Why Support Stacks Get Expensive: Intercom, Zendesk and Slack

A customer asks a question in the in-app chat. It gets escalated by email, which opens a ticket in a different system. The agent cannot answer it, so they paste it into a Slack channel where an engineer replies in eleven minutes. The customer gets a good answer. The company now has three partial records of one conversation, and the actual answer is in the one place that will never be searched again.

This stack is unusual, and the thing that makes it unusual is the pricing rather than the products. Most stacks mix per-seat and flat lines, so part of the bill stays still as you grow.

Here, every line is per seat. There is nothing in this combination that does not get more expensive when you hire, which makes support one of the few functions where the software bill tracks headcount almost exactly.

Entry-tier pricing from the registry our comparison pages use, at five people.

The number worth internalising is the middle one. $136 a month, per person, before they answer anything. That is roughly $1632 a year of software attached to each support hire, and it is never in the business case for the hire because it does not arrive as a purchase.

It is worth sitting with the incentive this creates, because it produces decisions that look like cost discipline and are not.

Teams ration seats. Someone who occasionally needs to look at a ticket does not get a licence, so they ask an agent to check for them. The cost moved from a subscription to an interruption, and interruptions are more expensive and invisible.

Escalation moves to where it is free. Engineers and product people do not have support seats, so support goes to them in Slack. Which is the next section, and it is the most expensive consequence of the pricing model.

Nobody removes a tool because everyone has both open. Once agents are used to two systems, consolidating means retraining during a period when tickets do not stop.

Slack is on this list because for most teams it is genuinely part of the support stack, even though nobody would describe it that way in a tools audit.

It is superb at the thing it is used for: getting a hard answer out of a person who knows it, fast. The problem is what happens to that answer. It stays in a thread. It does not reach the help centre, it is not attached to the customer record, the next agent with the same question cannot find it, and in ninety days it may not be retained at all.

The fix is not banning Slack, which will not work and should not. It is a deliberate step that takes the resolution back into the record, and the reason it does not happen is that it is unrewarded admin at the end of a resolved problem.

Very few teams sat down and chose to run both. It happens chronologically: Intercom for in-app chat when the product was young and the volume was small, Zendesk later when email support needed real ticketing, queues and SLAs.

Both products have since grown into each other. Intercom does ticketing, Zendesk does messaging. So a team that added the second one for a capability the first lacked is often now paying two per-seat bills for a heavily overlapping surface.

The reason it persists is not inertia so much as risk. Support is live: consolidating means moving conversation history, retraining agents and changing the customer-facing widget, all while tickets continue to arrive. It is a real project, which is why it keeps being scheduled for next quarter.

First, pick one conversation tool. Not because one is better, but because two per-seat bills for one job is the largest single saving available here and it compounds with every hire. Choose on which one your customers actually use, not on which the team prefers.

Second, audit who needs a full seat. Most teams find people paying full price to read occasionally. Some tools price light seats differently; the ones that do not are worth asking about at renewal.

Third, close the Slack loop. Anything resolved there needs one deliberate step back into the record. This saves no subscription and is probably worth more than either of the first two, because it stops you re-solving the same problem every quarter.

Do not migrate during your peak. Support has seasons too, and moving conversation history in the busiest month of the year is how consolidation gets a bad name.

The general migration sequence is in the GTM stack consolidation playbook, and the individual head-to-heads are on the Intercom, Zendesk and Slack comparisons. This post is about what the three cost together, which is the number that changes when you hire.