Hey everyone! I'm trying to get a handle on the support platform landscape for a future project. We're planning to build a new customer service setup in 2026, and omnichannel is a must-have.
I've been looking at Zendesk, Freshdesk, and Intercom basics, but all the marketing talks about "seamless" everything. I need something concrete. For those who have implemented these, what does the actual agent workflow look like when a customer switches from chat to email? Does the context actually carry over easily? Also, how do these platforms handle custom integrations? I'm used to thinking in AWS services, so I'm wondering if they have clear APIs or Terraform providers for managing parts of the setup.
Any real-world pros/cons on routing logic and reporting would be super helpful! 🙏
The context transfer is usually pretty good on the big platforms, but the real friction comes from agent notification and handoff. On a switch from chat to email, the agent might get a pop-up, an email, or just see the ticket updated in their queue. The delay there can feel clunky to a customer waiting for a reply.
If you're thinking AWS and Terraform, definitely dig into their REST APIs and webhook frameworks. Zendesk's API is mature, and Intercom is very developer-friendly for building custom workflows. I haven't seen dedicated Terraform providers, but you can manage a lot via their APIs and something like the AWS SDK.
On routing, the out-of-the-box rules are often too simple. You'll likely need to build custom logic based on customer data (like lifetime value or issue complexity) pulled from your own databases. That's where the integration work really pays off.
Let the machines do the grunt work
I'm also looking into this for our small team! You mentioned Zendesk and Freshdesk, which seem popular. I've heard that the context does carry over in the ticket history, but sometimes the agent has to manually open the chat transcript from the email view. It's not always automatic.
Since you're thinking about AWS and Terraform, have you looked at whether these platforms let you store chat logs or customer data directly in your own S3 buckets? I'm not technical enough to know about APIs, but our dev said something about webhooks being easier for custom alerts.
For routing, how do you plan to handle priority customers? Is that where the custom logic comes in?
Yeah, that manual step to open the chat transcript is a real workflow killer. I've seen it cause delays.
On storing data in your own S3 buckets, I know Zendesk's API can push events to webhooks. You could write a small Lambda function to receive those webhooks and archive the payloads to S3. That's usually easier than a direct integration. Webhooks are definitely simpler for custom alerts, like triggering a PagerDuty incident from a high-priority ticket.
For priority routing, custom logic is key. You'd need to pull customer tier from your own database via an API call to decide which queue to use.
Totally agree about the manual transcript step. I've seen that create a 2-3 minute lag where the customer thinks they've been dropped, and the agent is just hunting for context.
>write a small Lambda function
This is a solid approach. One caveat from my beta testing: watch out for event payload volume. If you're routing everything to S3, you might hit Lambda concurrency limits during a support spike. Setting up a dead-letter queue for those webhook events saved us.
For priority routing, pulling from your own DB is the way. We even used a similar API call to attach internal notes to the ticket automatically, so the agent knows *why* this customer got routed as high-priority.
edge cases matter