

TL;DR:
Industry benchmarks for ecommerce are hard to come by. Most of what's out there is self-reported, survey-based, or too aggregated to be usable. Teams are left wondering whether their AI adoption is on par with industry standards or if their response times are costing them revenue.
That's a gap we're in a unique position to close.
Gorgias processes millions of customer conversations across thousands of ecommerce brands every day. This has given us a rare, unfiltered view into how the industry operates. But until now, we’ve kept those insights largely internal.
Today, we're making it public with the Ecom Lab.
The result is years of first-party data from thousands of ecommerce brands, packaged into findings that give teams a real foundation to build their strategy on.
The Ecom Lab is Gorgias's public research hub for ecommerce. It publishes insights and reports on AI adoption, support performance, financial impact, and industry trends.
The goal is simple: give teams a real baseline to measure against and to uncover the industry's inner workings.
Metrics that actually move decisions.
The Ecom Lab publishes metrics that matter to ecommerce professionals, including AI adoption rates, first response times, CSAT scores, conversion rates, and ticket intents, all broken down by brand size, GMV tier, and industry vertical.
For the first time, teams can see exactly where they stand in comparison to the broader market.
AI is Everywhere reveals why roughly 4 in 5 ecommerce brands still haven't deployed AI in customer-facing support.
Stop Benchmarking Against the Average argues that support teams should benchmark response times against their specific industry vertical rather than the overall average.
Most Brands are Overpaying for Support breaks down the actual cost of support ticket volume and what happens when AI handles the load.
Four months ago, our analysts were dealing with a barrage of questions. "What's our ARR by segment?" "Build me a dashboard for this quarter's pipeline." Quick asks piled up behind complex deep dives. Stakeholders waited for answers that should have taken seconds, and analysts spent their time fielding requests instead of doing the strategic work that creates the most value.
Today, anyone at Gorgias can ask a question in plain language and get an accurate, contextualized response in seconds. Not from a colleague or dashboard, nor from a generic answer from the internet. But a response built on our business context. We call it Cortex, our flagship internal AI agent.
In two months, Cortex went from an idea to fielding thousands of questions every week, recommending actions across the business, and deprecating the need for manual dashboard creation. While most companies right now are treating AI as an initiative — at Gorgias, AI is already part of how we work. 72% of Gorgias employees use Cortex each week, and that number is only growing.
We didn’t achieve this by simply plugging a large language model into our stack. LLMs are a critical part of the equation, but they aren't the driving force — it’s everything else under the hood: the infrastructure, context, platform architecture, and the team that brings it all together.

The instinct across many companies today is to start with the model, pick a provider to solve a specific challenge, or invest heavily in getting the data right. All reasonable starting points, but most of them solve for one use case. Underneath that approach is a framing problem: seeing AI as an initiative — something you assign and measure. Seeing AI as another tool your company uses versus how your company operates.
We started somewhere different. Every company is built on four pillars: customers, people, product, and decisions. AI investments tend to place heavy emphasis on the first three. We started with the fourth. Our bet was that if we built everything around the need to make effective decisions first, asking what Gorgias needed to know to operate well, then our AI would become dramatically more powerful.
Cortex is our flagship internal AI agent, and the product where we established the tenets that now run through everything else we build: composable and modular infrastructure, governed context, and accessible from wherever decisions happen. Cortex lives in Slack, as well as across LLM vendors, in its own browser extension, and even on its own dedicated internal site.
Cortex doesn’t stop at answering questions. It can read and write to Notion, file Linear tasks, create HTML apps, automate signal delivery, and more. It operates across every layer of our stack, from dashboards to data pipelines, because we designed it as one integrated system. It is this connection that adds remarkable depth to what people can ask, and what they get in return.

A Sales Lead is pitching and asks Cortex for the full picture of the merchant. In a customized PDF, Cortex lists coverage gaps, pre-sale intent signals, and product fit options. Everything the sales lead needs to walk in with confidence.
A Senior Product leader asks, "How are we performing against OKR #1, and what can my team do to help accelerate it?" Cortex returns a full ARR breakdown, projected end-of-month attainment, segment-level findings, and connects it all back to company-level strategies. A suite of recommendations customized to the leader, the performance, and the signals that bridge how they can support our goals. The kind of answer that used to take someone a week to put together.
These aren't simple lookup queries. They require deep business context spanning multiple areas. Cortex handles these because its Decision Engine gives it the information to reason against governed data, metric definitions, and business context, turning a generic answer into a credible one.
Overnight, teams have built Cortex into how they work. They’re spending less time searching and more time finding answers, not because they were told to, but because Cortex reduced the distance between question and decision.
Cortex’s modular infrastructure allows us to experiment and add new capabilities freely. We’ve already built two more internal AI agents made for entirely different use cases, but using the same Decision Engine as Cortex.
GAIA, our internal experimentation AI Agent, helps our customers identify opportunities in their AI Agent Guidance design. It takes institutional knowledge across our teams and turns it into a scalable system that drives automation and value to our customers. Our CEO, Romain Lapeyre, has been its most vocal advocate since day one.
When we needed a platform for investor readiness and board preparation, we built Oracle. Our board decks and talk tracks are informed and built with the same AI, and our numbers are validated every step of the way.
We’re continuing to expand new AI agents internally, exploring how they can create value for customers and our own teams.
When AI handles thousands of analytical questions each week, the highest-value work for a data team shifts permanently. Late 2025, we repositioned from a Data Analytics function into a Decision Intelligence function — a structural change in what we own and how we operate.
Today, our analysts focus on the most sensitive, complex, and forward-looking decisions and analyses. They partner more deeply with stakeholders by driving next steps from signals. They're even building entirely new capabilities that didn't exist in their role descriptions months ago. Things like AI skills for Cortex, context curation, and insight and recommendation delivery. The role of the analyst hasn't diminished. It's expanded to encompass the most meaningful work an analyst can do: driving outcomes and ensuring those decisions can achieve them.

Our business support model has changed, too. Instead of embedding analysts and dedicated engineers within functional teams, we align capacity to the highest-impact company objectives and move fluidly across them. This model works even better because Decision Intelligence brings together both analytics and engineering teams under one roof.
Elliot Trabac leads our Data, Context and AI Engineering teams. The Decision Engine, Cortex, GAIA, and the platforms I've described exist because of the infrastructure his team innovated and built from the ground up. Noemie Happi Nono leads our Decision Strategy and Operations team, driving decision outcomes with stakeholders, advancing the development of Cortex skills and capabilities, and pushing into new areas of analysis every day.
Together, they're shaping what a modern data function looks like when AI becomes a standard building block for how a company operates.
The question of ROI is long gone. AI has opened the floodgates to more trusted and meaningful signals than ever. The natural next evolution is Proactive Intelligence, signals surfaced toward what you need to know, before you ask. And we're already building this because our architecture is designed to support it.
In the coming weeks, members of the Decision Intelligence team will go deeper into themes I've touched on here. Yochan Khoi, a Senior Analytics Engineer on our team, recently published a technical walkthrough of our context layer and will go further into building context strategies that scale. Others will cover infrastructure, analytical partnerships, evolving data assets into decision assets, and the cost and efficiency gains that make sustained AI investment viable.
AI hasn't changed the most important element of data and analytics functions — delivering outcomes — but it has raised the bar for what it looks like and how far we can take it. We’re just getting started.
The best in CX and ecommerce, right to your inbox

TL;DR:
The page-based shopping experience dominated for decades. Customers would search, browse, compare, abandon, get retargeted, return, and eventually buy (sometimes).
That journey is no longer the only option.
Shoppers are turning to chat, messaging, and AI-powered tools to find what they need. Instead of clicking through product pages or reading static FAQs, they ask questions, have back-and-forth conversations, and get answers that move them closer to a purchase in real time. The path to checkout has changed, and the brands that recognize this are pulling ahead.
Read our 2026 State of Conversational Commerce Report to learn more about conversation commerce trends from 400 ecommerce decision-makers and 16,000+ ecommerce brands using Gorgias.
{{lead-magnet-1}}
The traditional shopping journey was a solo experience. A shopper had a need, searched for options, browsed across sessions, and eventually made a decision — often days later, after being retargeted multiple times. Support only entered the picture after the purchase.

The conversation-led journey collapses that timeline:
What used to take days now takes minutes. Discovery, evaluation, and purchase happen in a single thread.
79% of brands agree that AI-driven conversational commerce has increased sales and purchase rates in their business. When brands were asked to rank the highest-return areas:
Those numbers reflect something important: the value of conversation compounds. Faster support reduces friction. Better retention raises lifetime value. More confident shoppers buy more often and spend more per order.
The brands seeing the biggest returns aren't just using AI to deflect tickets. They're using it to create one-to-one shopping experiences at scale.
Looking at AI-only influenced orders across key verticals like Apparel and Accessories, Food and Beverages, Health and Beauty, Home and Garden, and Sporting Goods, the growth across a single year was significant.





Across industries, ecommerce brands saw AI step into conversations, reduce shopper hesitation, and drive higher QoQ conversion rates.
Learn more about AI-powered revenue generation in the full 2026 Conversational Commerce Report.
84% of brands say the strategic importance of conversational commerce is higher than it was a year ago. 82% agree it will be mainstream in their sector within two years.

That shift is registering at the leadership level because of what conversational commerce does to the buying experience. Creating one-to-one touchpoints earlier in the journey drives higher AOV, shorter buying cycles, and stronger purchase rates. Shoppers who get real-time answers to their questions are more confident.
TUSHY, known for eco-friendly bidets and bathroom essentials, is a useful example of what happens when you take conversational commerce seriously.
Bidets aren't an impulse purchase. Shoppers have real questions about fit, compatibility, and installation. Those questions used to go unanswered until the CX team could respond, often after the customer had abandoned the cart.
TUSHY used Gorgias's AI Agent and shopping assistant capabilities to automate pre-sales support. AI Agent engaged shoppers in real-time conversations, addressed their concerns directly, and built confidence at the moment of highest intent.
This resulted in a 190% increase in chat-based purchases, a 13x return on investment, and twice the purchase rate of human agents.
You don't need to overhaul your entire operation to start seeing results. The most effective approach is to start where the impact is clearest and expand from there.
A few places to begin:
Want to see the full picture of where conversational commerce is headed in 2026? Read the full report to explore the data, trends, and strategies shaping the next era of ecommerce.
{{lead-magnet-1}}

TL;DR:
A year ago, ecommerce brands were still debating whether AI was worth the investment. That debate is over. Today, nearly every ecommerce professional uses AI to do their job.
The shift isn't just about adoption. It's about what AI is used for and how brands measure its impact. Support automation was the entry point. Now, AI is embedded across the full operation, from product recommendations to inventory control to real-time shopping conversations.
In our 2026 State of Conversational Commerce Report, we break down trends on AI usage among 400 ecommerce decision-makers and 16,000+ ecommerce brands using Gorgias.
{{lead-magnet-1}}
If we rewind 12 months ago, the industry was still split on AI. Some ecommerce professionals were excited, but most were still hesitant. In 2024, 69% of ecommerce professionals used AI in their roles. By 2025, that number reached 77%. In 2026, it hit 96%.

The confidence numbers back it up. 71% of brands say they are confident using AI for ecommerce, and 73% are satisfied with its business impact.
In early 2025, only 30% of ecommerce professionals rated their excitement for AI at 10/10. Today, zero percent of respondents describe themselves as hesitant about AI.

Using AI in ecommerce is not new. In fact, it dates back to the 1980s with the invention of algorithms and expert systems. And if you’ve ever leveraged similar product recommendations or chatbots, you’ve already integrated AI into your ecommerce stack.
Modern AI is far more sophisticated.
With the rise of agentic commerce and conversational AI, brands began leveraging AI agents to automate the processing of repetitive support tickets. That’s still happening today, but the scope has expanded beyond the support queue.

Ecommerce brands are deploying AI across every layer of their operation:
When brands were asked which channels contribute most to their AI success, conversational channels dominated. Social media messaging led at 78%, followed by SMS at 70%, and website live chat at 51%. Shoppers want fast, personal conversations, and AI is the best way to deliver that at scale.
Learn more about AI adoption, perception, and use case trends in the full 2026 Conversational Commerce Report.
For decades, customer support success meant fast response times and high satisfaction scores. Those are still important indicators of success, but leading brands are adding revenue-focused metrics to their dashboards.
91% of brands still track CSAT as a measure of AI's impact. But 60% now include AOV as a top indicator, and higher-revenue brands earning $20M+ are focusing on metrics like total operating expenses, cost per resolution, incremental revenue, and one-touch ticket rate.

AI can now start a conversation, ease customer doubts, sell, upsell, and recover abandoned carts in a single conversation. When you’re only measuring CSAT, you’re ignoring the real ROI of conversational AI investment.
Virtual shopping assistants now proactively engage shoppers, adapt to their needs in real time, and offer contextual product recommendations and upsells. When the moment calls for it, they can close the deal with a targeted discount.
Gorgias brands using AI Agent's shopping assistant capabilities nearly doubled their purchase rates and converted 20–50% better than those using AI Agent for support only.
Orthofeet, the largest provider of orthopedic footwear in the US, is a concrete example of this in practice. Using Gorgias, they achieved:
The data tells a clear story: AI has evolved beyond a tool for handling tier 1 support tickets. It’s a core part of your revenue generation strategy.
57% of brands are already using AI for 26–50% of all customer interactions, and 37% expect that share to rise to 51–75% within the next two years. The brands building toward that range now are the ones who will have the operational advantage when it matters most.
The practical question isn't whether to invest in AI. It's where to focus first. Based on where brands are seeing the most impact, three priorities stand out:
Want to go deeper on the full 2026 conversational commerce trends? Read the complete report for data across every major AI use case in ecommerce.
{{lead-magnet-1}}

TL;DR:
The way shoppers buy online has shifted and customers are at the center.
They no longer want to scroll through product pages, dig through FAQs, or wait 24 hours for an email reply. They open a conversation, ask a specific question, and expect a useful answer in seconds. Brands that can’t deliver these experiences at scale are seeing customer hesitation turn into abandoned carts and lost revenue.
This shift has a name: conversational commerce. It's the practice of using real-time, two-way conversations as your primary sales channel, through chat, AI agents, messaging apps, and voice.
What started as an experiment for early adopters has become a key growth lever, with 84% of ecommerce brands treating conversational commerce as a strategic pillar this year vs. last year.

We surveyed 400 ecommerce decision-makers across North America, the U.K., and Europe to understand how conversational commerce and AI are reshaping the ecommerce landscape. These findings are complemented by aggregated and anonymized internal Gorgias platform data from 16,000+ ecommerce brands.
The State of Conversational Commerce in 2026 trends report breaks down all of the findings, including five key trends shaping the ecommerce landscape.
{{lead-magnet-1}}
A few years ago, adding an AI chatbot to your site that could provide tracking links and Help Center article recommendations was a differentiator. Today, it's table stakes. McKinsey found that 71% of shoppers expect personalized experiences, and 76% get frustrated when they don't get them.
Right now, most ecommerce professionals use AI, with 93% having used it for at least 1 year. Enthusiasm is accelerating quickly, with only 30% of ecommerce professionals rating their excitement for AI at 10/10 in April 2025. Similarly, while AI adoption rose steadily year over year, it reached a clear peak in 2026.

The use cases driving this adoption are practical and high-volume:

These are the tickets that flood brands’ inboxes every day. AI agents resolve them instantly, without pulling teams away from conversations that actually require human judgment.
Explore AI adoption and use case data in more depth in the full report.
The traditional ecommerce funnel, visit site, browse products, add to cart, check out, is losing ground. Shoppers now discover products on Instagram, ask questions via direct message, and complete purchases without ever visiting a website.

Conversational AI is actively increasing revenue, with 79% of brands reporting that AI-driven interactions have increased sales and conversion in their business.

The practical implication is that every channel is becoming a storefront. Creating personalized touchpoints with customers earlier in the journey, through proactive engagement, is impacting the bottom line.
Read the full report to explore how AI conversions have increased QoQ by industry.
Pre-purchase hesitation is one of the biggest conversion killers in ecommerce. A shopper lands on your product page, has a question about sizing or compatibility, can't find the answer quickly, and leaves. That's a lost sale that had nothing to do with your product.
Conversational AI changes that dynamic. When a shopper can ask a question and get an accurate, personalized answer in real time, the friction disappears.
Brands using Gorgias saw this play out at scale in 2025. When AI Agent recommended a product, 80% of the resulting purchases happened the same day, and 13% happened the next day.

Brands are further accelerating the buying cycle through proactive engagement. On-site features such as suggested product questions, recommendations triggered by search results, and “Ask Anything” input bars drove 50% of conversation-driven purchases during BFCM 2025.
Explore how AI is collapsing the purchase cycle in Trend 3 of the report.
There's a persistent narrative that AI is making CX teams redundant. The data tells a different story. 62% of ecommerce brands are planning to grow their teams, not cut them. But the scope of those teams is changing.

New roles are emerging around AI configuration and quality assurance. Teams are investing in technical members to write AI Guidance instructions, develop tone-of-voice instructions, and continuously QA results.
CX teams are also bridging the gap between support goals and revenue goals, as the two functions increasingly overlap.

The result is CX teams that are more technical than they were before. Agents who once spent their days answering repetitive tickets are now spending that time on higher-value work: complex escalations, VIP customer relationships, and improving the AI systems and knowledge bases that handle the volume.
Learn more about the evolution of CX roles in Trend #4.
Despite increasing AI adoption, data shows that ecommerce brands shouldn’t strive for 100% automation. Winning brands are building systems in which AI handles repetitive tier-1 tickets, and humans handle complex, sensitive cases.

AI handles speed and scale. It resolves order-tracking requests at 2 a.m., processes return-eligibility checks in seconds, and answers the same shipping question for the thousandth time without compromising quality.
Human agents handle conversations that require context, empathy, or decisions that fall outside the standard playbook. There are several topics where shoppers still prefer human support.

Successful hybrid systems require continuous iteration, meaning reviewing handover topics, Guidance, and reviewing AI tickets on a weekly basis.
Discover how leading brands are balancing human and AI systems in Trend #5.
The 2026 trends are about expansion and standardization. The 2030 predictions are about what comes next.

Voice-based purchasing is the biggest bet on the horizon. Only 7% of brands currently use voice assistants for commerce, but 89% expect it to be standard by 2030. The vision is a customer who can reorder a product, check their subscription status, or manage a return entirely over the phone.
Proactive AI is the other major shift. Rather than waiting for a customer to reach out, AI will anticipate needs based on browsing behavior, purchase history, and where someone is in their relationship with your brand. Think of it as the digital equivalent of a sales associate who remembers what you bought last time and knows what you're likely to need next.
Explore where ecommerce brands are allocating their AI budgets in the full report.
The brands winning in 2026 are creating smart, scalable systems where AIhandles volume and humans handle nuance. They’re treating every conversational channel as an opportunity to serve and sell.
The data is clear: AI adoption is accelerating, customer expectations are rising, and the revenue impact of getting this right is measurable.
{{lead-magnet-1}}

Celery just released their API on Github, currently in beta. Here are some of the cool stuff you can do with it in Gorgias.
When you receive an email from a customer, you can connect your Celery account and see customer information (orders, shipping address, etc.). Here’s what it looks like:

To configure it, grab your Celery access_token, head to integrations, and add an HTTP integration using this URL:
https://api.trycelery.com/v2/orders?buyer.email={ticket.requester.email}
Then you can customize the sidebar to only show the Celery data you need to respond to customers. Click the cog and simply drag and drop elements you want to show.

Celery’s API enables you to perform a few actions from your favorite helpdesk:
Here’s an example of how you can cancel an order from Gorgias itself. Say you already have a macro to cancel an order. Add an HTTP action to it, in this case:
https://api.trycelery.com/v2/orders/{ticket.requester.customer.data[0].number}/order_cancel
Then, when you use this macro and send it to the customer, it will automatically cancel the last order at the same time:

We hope this integration with Celery can save you time. If you'd like to try Celery with Gorgias, shoot us a note! At support@gorgias.com.

TLDR: https://github.com/xarg/pghoard-k8s
This is a small tutorial on how to do incremental backups using pghoard for your PostgreSQL (I assume you’re running everything in Kubernetes). This is intended to help people to get started faster and not waste time finding the right dependencies, etc..
pghoard is a PostgreSQL backup daemon that incrementally backups your files on a object storage (S3, Google Cloud Storage, etc..).
For this tutorial what we’re trying to achieve is to upload our PostgreSQL to S3.
First, let’s create our docker image (we’re using the alpine:3.4 image cause it’s small):
FROM alpine:3.4
ENV REPLICA_USER "replica"
ENV REPLICA_PASSWORD "replica"
RUN apk add --no-cache \
bash \
build-base \
python3 \
python3-dev \
ca-certificates \
postgresql \
postgresql-dev \
libffi-dev \
snappy-dev
RUN python3 -m ensurepip && \
rm -r /usr/lib/python*/ensurepip && \
pip3 install --upgrade pip setuptools && \
rm -r /root/.cache && \
pip3 install boto pghoard
COPY pghoard.json /pghoard.json.template
COPY pghoard.sh /
CMD /pghoard.sh
REPLICA_USER and REPLICA_PASSWORD env vars will be replaced later in your Kubernetes conf by whatever your config is in production, I use those values to test locally using docker-compose.
The config pghoard.json which tells where to get your data from and where to upload it and how:
{
"backup_location": "/data",
"backup_sites": {
"default": {
"active_backup_mode": "pg_receivexlog",
"basebackup_count": 2,
"basebackup_interval_hours": 24,
"nodes": [
{
"host": "YOUR-PG-HOST",
"port": 5432,
"user": "replica",
"password": "replica",
"application_name": "pghoard"
}
],
"object_storage": {
"aws_access_key_id": "REPLACE",
"aws_secret_access_key": "REPLACE",
"bucket_name": "REPLACE",
"region": "us-east-1",
"storage_type": "s3"
},
"pg_bin_directory": "/usr/bin"
}
},
"http_address": "127.0.0.1",
"http_port": 16000,
"log_level": "INFO",
"syslog": false,
"syslog_address": "/dev/log",
"syslog_facility": "local2"
}
Obviously replace the values above with your own. And read pghoard docs for more config explanation.
Note: Make sure you have enough space in your /data; use a Google Persistent Volume if you DB is very big.
Launch script which does 2 things:
#!/usr/bin/env bash
set -e
if [ -n "$TESTING" ]; then
echo "Not running backup when testing"
exit 0
fi
cat /pghoard.json.template | sed "s/\"password\": \"replica\"/\"password\": \"${REPLICA_PASSWORD}\"/" | sed "s/\"user\": \"replica\"/\"password\": \"${REPLICA_USER}\"/" > /pghoard.json
pghoard --config /pghoard.json
Once you build and upload your image to gcr.io you’ll need a replication controller to start your pghoard daemon pod:
apiVersion: v1
kind: ReplicationController
metadata:
name: pghoard
spec:
replicas: 1
selector:
app: pghoard
template:
metadata:
labels:
app: pghoard
spec:
containers:
- name: pghoard
env:
- name: REPLICA_USER
value: "replicant"
- name: REPLICA_PASSWORD
value: "The tortoise lays on its back, its belly baking in the hot sun, beating its legs trying to turn itself over. But it can't. Not with out your help. But you're not helping."
image: gcr.io/your-project/pghoard:latest
The reason I use a replication controller is because I want the pod to restart if it fails, if a simple pod is used it will stay dead and you’ll not have backups.
Future to do:
Hope it helps, stay safe and sleep well at night.
Again, repo with the above: https://github.com/xarg/pghoard-k8s

At Gorgias we recently switched our flask & celery apps from Google Cloud VMs provisioned with Fabric to using docker with kubernetes (k8s). This is a post about our experience doing this.
Note: I'm assuming that you're somewhat familiar with Docker.
The killer feature of Docker for us is that it allows us to make layered binary images of our app. What this means is that you can start with a minimal base image, then make a python image on top of that, then an app image on top of the python one, etc..
Here's the hierarchy of our docker images:
Piece of advice: If you used to run your app using supervisord before I would advise to avoid the temptation to do the same with docker, just let your container crash and let k8s handle it.
Now we can run the above images using: docker-compose, docker-swarm, k8s, Mesos, etc...
There is an excellent post about the differences between container deployments which also settles for k8s.
I'll also just assume that you already did your homework and you plan to use k8s. But just to put more data out there:
Main reason: We are using Google Cloud already and it provides a ready to use Kubernetes cluster on their cloud.
This is huge as we don't have to manage the k8s cluster and can focus on deploying our apps to production instead.
Let's begin by making a list of what we need to run our app in production:
We ran the above in a normal VM environment, why would we need k8s? To understand this, let's dig a bit into what k8s offers:
There are more concepts like volumes, claims, secrets, but let's not worry about them for now.
We're using Postgres as our main storage and we are not running it using Kubernetes.
Now we are running postgres in k8s (1 hot standby + pghoard), you can ignore the rest of this paragaph.
The reason here is that we wanted to run Postgres using provisioned SSD + high memory instances. We could have created a cluster just for postgres with these types of machines, but it seemed like an overkill.
The philosophy of k8s is that you should design your cluster with the thought that pods/nodes of your cluster are just gonna die randomly. I haven't figured our how to setup Postgres with this constraint in mind. So we're just running it replicated with a hot-standby and doing backups with wall-e for now. If you want to try it with k8s there is a guide here. And make sure you tell us about it.
RabbitMQ (used as message broker for Celery) is running on k8s as it's easier (than Postgres) to make a cluster. Not gonna dive into the details. It's using a replication controller to run 3 pods containing rabbitmq instances. This guide helped: https://www.rabbitmq.com/clustering.html
As I mentioned before, we're using a replication controller to run 3 pods, each containing uWSGI & NGINX containers duo: gorgias/web & gorgias/nginx. Here's our replication controller web-rc.yaml config:
apiVersion: v1
kind: ReplicationController
metadata:
name: web
spec:
replicas: 3 # how many copies of the template below we need to run
selector:
app: web
template:
metadata:
labels:
app: web
spec:
containers:
- name: web
image: gcr.io/your-project/web:latest # the image that you pushed to Google Container Registry using gcloud docker push
ports: # these are the exposed ports of your Pods that are later used by the k8s Service
- containerPort: 3033
name: "uwsgi"
- containerPort: 9099
name: "stats"
- name: nginx
image: gcr.io/your-project/nginx:latest
ports:
- containerPort: 8000
name: "http"
- containerPort: 4430
name: "https"
volumeMounts: # this holds our SSL keys to be used with nginx. I haven't found a way to use the http load balancer of google with k8s.
- name: "secrets"
mountPath: "/path/to/secrets"
readOnly: true
volumes:
- name: "secrets"
secret:
secretName: "ssl-secret"
And now the web-service.yaml:apiVersion: v1
kind: Service
metadata:
name: web
spec:
ports:
- port: 80
targetPort: 8000
name: "http"
protocol: TCP
- port: 443
targetPort: 4430
name: "https"
protocol: TCP
selector:
app: web
type: LoadBalancer
That type: LoadBalancer at the end is super important because it tells k8s to request a public IP and route the network to the Pods with the selector=app:web.
If you're doing a rolling-update or just restarting your pods, you don't have to change the service. It will look for pods matching those labels.
Also a replication controller that runs 4 pods containing a single container: gorgias/worker, but doesn't need a service as it only consumes stuff. Here's our worker-rc.yaml:
apiVersion: v1
kind: ReplicationController
metadata:
name: worker
spec:
replicas: 2
selector:
app: worker
template:
metadata:
labels:
app: worker
spec:
containers:
- name: worker
image: gcr.io/your-project/worker:latest
With Kubernetes, docker finally started to make sense to me. It's great because it provides great tools out of the box for doing web app deployment. Replication controllers, Services (with LoadBalancer included), Persistent Volumes, internal DNS. It should have all you need to make a resilient web app fast.
At Gorgias we're building a next generation helpdesk that allows responding 2x faster to common customer requests and having a fast and reliable infrastructure is crucial to achieve our goals.
If you're interested in working with this kind of stuff (especially to improve it): we're hiring!

We've released a new version of the Chrome Extension, with sharing features and a new navigation bar. We hope you'll love it!
Before, the only way to share templates with your teammates was to login on Gorgias.io.
If you're on the startup plan, when you create a template, you can choose who has access to it: either only you, specific people, or your entire team.

The account management section is now available in the extension, under settings.
Tags are now available on the left. It's easier to manage hundreds of templates with them.
You can also navigate through your private & shared templates. Shared templates include templates shared with specific people or with everyone.

We hope you'll enjoy this new version of our Chrome Extension. As usual, your feedback & questions are welcome!

Today, we’re thrilled to announce that we’ve raised a $1.5 million Seed round led by Charles River Ventures and Amplify Partners, to help build our new helpdesk.
We’re incredibly grateful to early users, customers, mentors we’ve met both at and Techstars.
We started the journey with Alex at the beginning of 2015 with our Chrome extension, which helps write email faster using templates. We’ve been pleased all along with customers telling us about how helpful it was, especially for customer support.
While building the extension, we’ve realized that a big inefficiency in support lies in the lack of integration between the helpdesk, the payment system, CRM and other tools support is using. As a result, agents need to do a lot of repetitive work to respond to customer requests, especially when the company is big.

That’s why we’ve decided to build a new kind of helpdesk to enable customer support agents to respond 2x faster to customers. You can find out more and sign up for our private beta here.
When a company has a lot of customers, support becomes repetitive. We want to provide support teams with tools to automate the way they treat simple repetitive requests. This way, they have more time for complex customer issues.
We'll now focus on this helpdesk and on growing the team, oh, and if you'd like to join, we're hiring! We're super excited about this new helpdesk product. If you’re using the extension, don’t worry.
Romain & Alex

Last few months we got lots of feedback about our extension and found to our delight that most people are satisfied, but still a few recurrent issues came up:
We listened and now we're presenting:
WYSIWYG editors for the web are notoriously buggy and are just difficult to develop.
I have yet to see one that is bug free. There are few venerable editors that do a good job like TinyMCE, FKEditor or CKEditor.. but they are big and all have edge cases that break the intended formatting and add a lot of garbage html.
There are newer good quality editors in town such as Redactor. The one that got my attention and finally landed in Gorgias is this wonderful editor called which is super lightweight, uses modern content-editable (no i-frames) and 'just works' most of the time. That's not to say it's perfect, but it's good enough and I'm satisfied with it's direction in terms of development.
Enjoy it and as always send us bug-reports or feedback on: support@gorgias.com

TL;DR:
A CX helpdesk is the command center where ecommerce brands manage every customer conversation, from pre-purchase questions to post-purchase support.
Unlike traditional IT helpdesks that focus on internal tickets, CX helpdesks are built for customer-facing teams who need to resolve issues fast while driving sales. Modern platforms combine ticket management with AI automation, letting you handle support at scale without sacrificing quality.
For ecommerce brands specifically, the right CX helpdesk turns support from a cost center into a revenue driver by enabling agents to recommend products, process returns, and recover abandoned carts — all from one screen.
A customer experience (CX) helpdesk is a software platform that unifies all customer conversations from every channel into a single, shared inbox. This means emails, live chat messages, social media comments, and SMS texts all land in one place, giving your support team a complete view of every customer interaction.
CX stands for customer experience, which is the entirety of a customer's interaction with your brand. This includes everything from browsing your website to receiving their order and getting help when they need it. A CX helpdesk focuses specifically on managing these conversations to create positive experiences that keep customers coming back.
Unlike a generic helpdesk built for IT or B2B companies, a CX helpdesk for ecommerce is designed around the shopper. It integrates directly with your store backend, like Shopify, to pull in crucial data like order history, shipping status, and past purchases. This allows agents to resolve issues without switching tabs, turning a simple question into a fast, personalized experience.
The core of any CX helpdesk is its ticketing system. A ticketing system is a tool that organizes each customer inquiry into a trackable ticket with a unique ID, status, and history. From there, automation and AI can tag, prioritize, and even resolve common questions automatically.
Key components of a CX helpdesk include:
Choosing the right platform depends on your store's scale, your team's needs, and your primary goals. We evaluated the top options based on their ecommerce integrations, automation capabilities, and ability to drive revenue, not just resolve tickets.
Gorgias is a conversational commerce platform built specifically for ecommerce brands. It combines a powerful helpdesk with an AI Agent designed to both resolve support tickets and convert shoppers into customers. Its deep, two-way integration with Shopify is a core differentiator, allowing teams to manage orders, issue refunds, and create discount codes directly within the helpdesk.
The platform's AI Agent can automate up to 60% of common inquiries like "Where is my order?" process returns, and answer product questions with brand-aligned responses. Beyond support, Gorgias includes revenue-driving tools that proactively engage shoppers with personalized recommendations and targeted chat campaigns.
What makes Gorgias different is its focus on conversational commerce. This means using real-time conversations as your storefront, where support and sales happen in the same place. Your team can recommend products, process orders, and resolve issues without customers ever leaving the chat.
Main features:
Ideal for: Shopify brands doing $1M+ USD in revenue, teams managing 500+ tickets monthly, brands prioritizing sales alongside support
Pricing: Starter: $10 USD/month (50 tickets), Basic: $60 USD/month (300 tickets), Pro: $360 USD/month (2,000 tickets), Advanced: $900 USD/month (5,000 tickets), Enterprise: Custom pricing
Read more: Which Gorgias plan should you choose? (Pricing breakdown)
Zendesk is one of the largest players in the customer service software market, offering a comprehensive suite of tools for businesses of all sizes. Its platform is highly customizable and scalable, making it a popular choice for large enterprises with complex support needs.
While not built exclusively for ecommerce, Zendesk offers robust integrations with Shopify and other platforms. Its strengths lie in its powerful ticketing system, extensive reporting capabilities, and mature AI features that can suggest answers and automate workflows.
However, its general-purpose nature means that achieving the deep ecommerce functionality native to platforms like Gorgias often requires more setup, customization, and reliance on third-party apps. You'll need technical expertise to get the most out of Zendesk's advanced features.
Ideal for: Large enterprises with dedicated IT teams, businesses serving multiple industries, companies needing extensive customization
Pricing: Suite Team: $55 USD/agent/month, Suite Growth: $89 USD/agent/month, Suite Professional: $115 USD/agent/month, Suite Enterprise: Custom pricing
Read more: Zendesk pricing in 2026: Plans, add-ons, and if it’s worth it
Freshdesk is known for its user-friendly interface and affordable pricing, making it an attractive option for small and medium-sized businesses. It provides a solid set of helpdesk features, including omnichannel ticketing, automation, and a self-service knowledge base.
Its "Freddy AI" offers chatbot capabilities and agent assistance. Like Zendesk, Freshdesk serves a wide range of industries, so its ecommerce features are not as deeply embedded as in specialized platforms. While it integrates with Shopify, the level of direct order management within the helpdesk is more limited.
Freshdesk works well for brands looking for a straightforward and cost-effective helpdesk solution that covers the basics without overwhelming complexity.
Ideal for: Small to medium businesses, teams new to helpdesk software, brands with straightforward support needs
Pricing: Free: $0 USD (up to 10 agents), Growth: $15 USD/agent/month, Pro: $49 USD/agent/month, Enterprise: $79 USD/agent/month
Read more: Freshdesk pricing guide: What you actually pay in 2026
Intercom positions itself as a "customer communications platform" with a strong focus on proactive engagement through its live chat and chatbot products. It excels at engaging website visitors, qualifying leads, and onboarding new users, making it popular with SaaS companies and brands with a heavy focus on marketing-led conversations.
For ecommerce, Intercom's strength is in its pre-purchase engagement capabilities. Its chatbots can guide shoppers, recommend products, and capture leads effectively. However, its post-purchase support functionality and deep backend ecommerce integrations are less developed compared to platforms built specifically for support-heavy retail operations.
Ideal for: SaaS companies, businesses focused on lead generation, brands prioritizing pre-purchase engagement
Pricing: Essential: $39 USD/seat/month, Advanced: $99 USD/seat/month, Expert: $139 USD/seat/month
A modern CX helpdesk streamlines the entire support process into a clear, repeatable workflow. It starts the moment a customer reaches out and ends with a fast, satisfying resolution that strengthens their relationship with your brand.
Here's how the process typically works. A customer sends a message via email, live chat, or a social media DM. The message automatically creates a ticket in a single, shared inbox, so nothing gets lost.
Smart routing happens next. Automation Rules instantly analyze the ticket based on the channel, keywords, or customer history. The ticket gets tagged (like "Return Request") and assigned to the correct agent or team.
When the agent opens the ticket, they immediately see the shopper's complete history in a side panel. This includes past orders from Shopify, previous conversations, and data from other integrated apps. Full context without switching tabs.
For direct actions, if the customer wants to make a return, the agent can process it directly within the helpdesk using an integration. No need to log into Shopify or a separate returns app.
Automation rules handle simple questions automatically. For a question like "Where is my order?" an automated workflow can pull the tracking information from Shopify and send an immediate, personalized reply, resolving the ticket without any agent involvement.
The key workflow components include:
Investing in a CX helpdesk isn't just about managing support tickets more efficiently. It's a strategic decision that directly impacts customer loyalty, operational costs, and your bottom line. When support is fast, personal, and helpful, it becomes a key driver of growth.
Retention and lifetime value: Fast, effective support is a major factor in customer retention. Solving a problem quickly and personally makes customers feel valued, encouraging them to make repeat purchases and increasing their lifetime value.
Conversion optimization: A CX helpdesk with proactive chat can engage hesitant shoppers on product or checkout pages. Answering a quick question about sizing or shipping can be the final nudge a customer needs to complete their purchase.
Operational efficiency: By automating repetitive questions and centralizing tools, a helpdesk dramatically reduces the time agents spend on manual tasks. This lowers your cost per ticket and allows a lean team to handle a high volume of conversations.
Revenue generation: Top CX helpdesks empower agents to be salespeople. With full customer context, they can recommend relevant products, upsell accessories, and even create draft orders directly in the chat, turning a support interaction into a sale.
The business impact extends beyond individual transactions. When customers know they can get help quickly, they're more likely to try new products, make larger orders, and recommend your brand to others. Support becomes a competitive advantage rather than just a cost center.
Modern ecommerce customers expect instant responses and personalized service. A CX helpdesk helps you meet these expectations consistently, even as your business grows. The alternative — missed messages, long response times, and frustrated customers — directly hurts your revenue and reputation.
A CX helpdesk is no longer just a tool for managing complaints. For modern ecommerce brands, it's the engine for building customer relationships, driving revenue, and scaling operations efficiently. By centralizing conversations and automating workflows, you can deliver the fast, personal experiences that today's shoppers expect.
The best time to implement a CX helpdesk is before you need it. Waiting until you're overwhelmed with support requests means you're already losing customers and revenue. Start with a platform that can grow with your business and adapt to your specific needs.
Ready to see how a platform built for ecommerce can turn your support center into a profit center? Book a demo to learn how you can transform every customer conversation into a growth opportunity.
{{lead-magnet-2}}

TL;DR:
Richpanel uses a dual pricing model that combines seat-based help desk costs with order volume-based portal fees. The platform starts at $29 per agent per month, but real costs quickly climb when you add the self-service portal, AI features, and channel integrations. Understanding how each component affects your total spend helps you budget accurately and compare alternatives. This guide breaks down exactly what you'll pay based on your team size, order volume, and feature needs.
Richpanel's pricing is built on two separate components that you pay for independently. You pay a per-agent fee for the help desk platform and a separate order-based fee for the self-service portal.
This means your total monthly cost isn't just the number of agents multiplied by the plan price. You must also factor in the cost of the customer-facing portal, which is required for self-service features and scales with your store's monthly order volume.
Seat-based pricing is a model where you pay a fixed monthly fee for each user or agent who has access to the software. This means if you have five agents, you pay five times the monthly rate. Order-based pricing ties costs directly to the number of orders your store processes each month.
The base help desk plans include core features, but many functions that modern ecommerce brands consider essential come at an additional cost:
Richpanel offers four main help desk tiers, each priced per agent per month. The plan you choose determines what features your support team can access, with significant jumps in functionality and cost between each level.
The Starter plan costs $29 per seat per month. It provides the fundamental tools for managing customer conversations, including a unified ticketing inbox, live chat, and basic automation Rules.
This plan is designed for very small teams or new stores just starting to centralize support. However, it lacks key efficiency features like collision detection, which prevents multiple agents from working on the same ticket simultaneously.
At $49 per seat per month, the Regular plan is the most common starting point for growing brands. It adds critical features like collision detection, advanced automation capabilities, custom ticket fields, and team collaboration tools.
This tier provides the necessary foundation for a team to handle a moderate volume of inquiries efficiently. Most brands find this plan meets their needs once they have more than two agents handling support.
The Pro plan costs $99 per seat per month and is built for larger teams that require deeper insights and control. It unlocks advanced customer interaction analytics, custom user roles and permissions, and API access for building custom integrations.
Brands typically upgrade to Pro when they need to manage multiple teams, enforce stricter access controls, or integrate support data with other business intelligence tools. The price jump from Regular to Pro is significant, so you'll want to ensure you actually need these advanced features.
The Enterprise plan uses custom pricing and is tailored for high-volume businesses with complex operational needs. It includes everything in Pro, plus features like single sign-on (SSO), a dedicated account manager, custom integration support, and service level agreement (SLA) guarantees.
This plan usually has a minimum seat requirement and is negotiated directly with Richpanel's sales team. Expect to pay significantly more than the Pro plan, often starting at several thousand dollars per month.
The self-service portal is a separate product with its own pricing structure tied to your monthly order volume. This is an additional cost on top of your help desk seat licenses.
The portal allows your shoppers to track orders, initiate returns, and find answers in a knowledge base without contacting an agent. This means you're essentially paying for two products: the agent-facing help desk and the customer-facing portal.
For stores processing up to 5,000 orders per month, the portal costs between $9 and $29 per month, depending on your help desk plan. This tier includes essential self-service features like order tracking, basic returns management, and a knowledge base.
The portal requires a direct sync with your Shopify store to function properly. Without this integration, customers can't access their order information or complete self-service actions.
As your order volume grows, so does the portal cost. The Pro portal plan is for stores with up to 20,000 monthly orders and costs between $49 and $99 per month.
It adds advanced features like deflection analytics to measure self-service success, custom branding for the portal, and multilingual support. The deflection analytics help you understand how many tickets the portal is preventing, which can justify the additional cost.
For brands exceeding 20,000 orders per month, the portal pricing is custom and negotiated as part of an Enterprise package. This tier offers volume-based discounts and advanced capabilities tailored to high-scale operations.
The exact cost depends on your order volume, but expect to pay several hundred dollars per month for portal access at this level.
Beyond the help desk and portal, several add-ons can significantly increase your total monthly spend. Many of these aren't optional extras but necessary components for running a comprehensive support operation.
Sidekick AI, Richpanel's AI assistant, costs an additional $10 per seat per month. It provides agents with suggested responses, sentiment detection, and auto-tagging for tickets.
Because it's priced per seat, this cost multiplies across your entire team. For a team of five agents, this adds $50 to your monthly bill on top of your base plan costs.
The social media moderator is a separate AI tool for managing Instagram and Facebook comments. It costs a flat fee of $49 per month for the entire account, regardless of how many agents you have.
This tool helps filter spam, hide negative comments, and create tickets from relevant social media interactions. Unlike Sidekick AI, this is an account-wide fee rather than a per-seat charge.
For teams that need help with setup, Richpanel offers an "Automation Success Kit" for a one-time fee of $499. This package includes guided onboarding, workflow templates, and training sessions to help you get the most out of the platform's automation features.
While this is optional, many brands find they need this support to properly configure their workflows and automation Rules.
Integrating channels like WhatsApp or voice isn't native to Richpanel and requires using third-party providers. This introduces variable and often unpredictable costs that can significantly impact your monthly spend.
WhatsApp integration relies on the WhatsApp Business API, which has its own fees based on message volume. Voice support requires a separate telephony provider with its own monthly fees and per-minute charges. These integrations can add anywhere from $50 to several hundred dollars to your monthly expenses, depending on usage.
Your total cost for Richpanel is driven by three main factors that compound: the number of agent seats, your monthly order volume, and the channels you support. A small change in one area can have a significant impact on your overall bill.
The largest and most direct cost driver is the number of agent seats. Each agent you add increases your monthly bill by the price of your chosen help desk plan.
This linear scaling means you need to be mindful of team size, especially if you rely on seasonal or part-time agents who may only need access for a few months a year. Unlike ticket-based pricing models, you pay for seats whether they're actively used or not.
Consider these monthly costs for different team sizes:
Your self-service portal fees are tied directly to order volume, which can create cost spikes during peak seasons like Black Friday Cyber Monday (BFCM). A brand that typically processes 4,000 orders a month might suddenly jump to 15,000 orders in November, pushing them into a higher, more expensive portal tier for that month.
This creates unpredictable costs during your busiest sales periods, when you're already dealing with increased support volume and potentially higher staffing costs.
Each additional support channel, especially voice and WhatsApp, adds another layer of cost and complexity. These often require separate subscriptions with third-party providers, making it difficult to forecast your total monthly spend.
Managing multiple vendors for core support functions also increases administrative overhead and can create gaps in your customer experience if integrations break or providers change their pricing.
When comparing Richpanel to other platforms, the most significant difference is the pricing model. Richpanel uses a traditional per-seat model, while a Shopify-native ecommerce helpdesk like Gorgias uses a ticket-based model.
A ticket-based pricing model means you pay based on the number of customer conversations you handle, not the number of agents on your team. This offers more flexibility, especially for businesses with fluctuating support volumes or seasonal patterns.
Here's how the models compare:
For brands with high seasonality, a ticket-based model like Gorgias often provides a lower total cost of ownership. You only pay for the support you actually provide, so your costs naturally scale down during slower months.
In contrast, a seat-based model requires you to pay for agent licenses year-round, even if they're not fully utilized during off-peak periods.
The help desk migration process from Richpanel to Gorgias is streamlined, with dedicated support to ensure a smooth transition of data and workflows. Most migrations take two to four weeks and include transferring your ticket history, macros, and automation Rules.
The customer support landscape offers several alternatives, each with a different approach to pricing and features. Understanding these options helps you make an informed decision about which model works best for your business.
Zendesk uses a per-seat model similar to Richpanel but with more complex tiers and add-ons.
Freshdesk also uses a per-seat model with various tiers. It offers a free plan with limited features, but costs can escalate quickly as you add agents and functionality. The free plan is often too restrictive for growing ecommerce brands.
Re:amaze is a more direct competitor in the ecommerce space and uses a per-user model. It bundles many features that Richpanel charges extra for, but can become costly for larger teams due to its per-user pricing structure.
When evaluating alternatives, look beyond the sticker price and consider the total cost of ownership:
Choosing a help desk is a major operational decision that affects your team's efficiency and your customers' experience. The right platform aligns with your budget, scales with your growth, and empowers your team to drive revenue rather than just resolve tickets.
Gorgias offers transparent, ticket-based pricing with no hidden fees for channels or users. Your costs are always tied to business activity, not arbitrary seat counts. Our AI Agent can reduce your support costs while increasing sales through intelligent automation and shopping assistance.
To see exactly what you would spend and how our platform can transform your customer experience, book a demo with one of our specialists.
{{lead-magnet-2}}


