Feature Bloat Is Costing You Sales: A 2026 Live-Chat Checklist for UK Teams of 2–20
Photo by Photo by Vitaly Gariev on Unsplash on Unsplash
Every live-chat demo runs to the same script: forty minutes of screens you will never open, ten minutes of pricing deflection, then a follow-up email with a PDF nobody reads. For a UK sales team of two to twenty people, the real decision is much narrower than the sales deck implies.
What follows is the checklist we would hand a sales lead who has three platforms shortlisted and a week to pick one. It assumes a few hundred conversations a month, one or two people on the chat rota at any moment, and a CRM that somebody occasionally remembers to update.
Start with cost per resolved conversation, not cost per seat
Vendors price per seat, per conversation, per resolution, or some combination that only becomes legible after a call with their finance team. Send every supplier the same written request: the total cost over twelve months at your expected volume, including the add-ons you would realistically need by month three.
If a quote cannot be produced in writing within two working days, that is your first data point. A supplier that hides the number during procurement tends to hide it on the renewal invoice too.
The twelve features that actually earn their keep
1. Skill and intent-based routing. Language, product line, or queue triage. With three agents this looks like overkill; with eight it is the difference between a sensible rota and one person drowning in refunds.
2. Genuine two-way CRM sync. Not a form dump into a lead record — field-level write-back, so an agent can update a deal stage from the chat window and see last month's order history without opening another tab.
3. A mobile agent app that was built on purpose. In a small team, the owner answers chats on a train. Test the app on a phone before you sign, on a real 4G connection in a real town.
4. Offline capture with a proper handover. Out-of-hours chat should collect an email, a reason for contact, and enough context that the morning reply is not a blank 'how can we help?'.
5. Canned responses with variables and shortcuts. Speed matters more than poetry. A saved answer that pulls in the customer's name and order number saves seconds on every single conversation.
6. Searchable transcripts kept for at least twelve months. Your best source of objection data is last quarter's chats. If the export is a CSV with broken line breaks, it is not searchable.
7. A workable UK GDPR toolkit. Consent captured at the point of chat, retention rules you can actually configure, a signed data processing agreement, and a straight answer about where data sits. Ask specifically about sub-processors and whether anything leaves the UK or EEA.
8. Transparent overage rules. Know what an extra thousand conversations costs before you have them, not after.
9. An API and webhooks. One connector to your helpdesk or order system will pay for itself faster than any bundled gimmick.
10. Proactive triggers. Fire an opener on cart value, page type, referrer, or time on page. This is where chat stops being a helpdesk and starts being a sales channel.
11. Agent-level reporting with revenue attribution. Response time, concurrent chats, and — crucially — which conversations ended in an order. If the report cannot connect a chat to a basket, you cannot defend the budget.
12. SSO, two-factor, and role permissions. Boring, essential, and increasingly a procurement requirement rather than a nice-to-have.
The five you can probably skip
- AI conversation summaries. Pleasant on a demo, rarely read by anyone who was not already on the chat.
- Gamified agent leaderboards. Badges do not fix a rota, and they encourage agents to close chats rather than solve them.
- Voice and video calling inside chat. Almost nobody wants a video call with a retailer at 9pm on a Sunday.
- Co-browsing. Genuinely useful in a handful of complex B2B sales cycles; not worth a line item for a team selling £40 products.
- Thirty-language auto-translation. Unless you ship internationally, it is a box-tick that adds contract complexity.
None of these are bad. They are simply the features that inflate a quote without touching the number your finance director cares about.
Score it in an afternoon, not a fortnight
Use the same weighted sheet for every supplier, and score what you have seen rather than what you have been told. Anything unproven scores zero until the trial.
| Criterion | Weight | Score (0–5) |
|---|---|---|
| Total year-one cost clarity | 20% | |
| CRM sync depth | 15% | |
| Routing and queue logic | 15% | |
| Mobile agent experience | 10% | |
| GDPR and data residency answers | 15% | |
| Proactive targeting | 10% | |
| Reporting and revenue attribution | 10% | |
| Migration and onboarding support | 5% |
A platform that scores badly on cost clarity and GDPR but wonderfully on routing is telling you something. So is the reverse.
The questions that end the call early
Ask how the free trial ends — does it auto-convert to a paid annual plan? Ask what happens to your transcripts if you leave, and in what format. Ask who actually supports you during December, and whether that is a UK number or a ticket queue on Pacific time. Ask for two references from businesses of a similar headcount, then ring them.
Vendors rarely fail these questions outright. They reveal themselves in how long the answers take to arrive.
The short version
For a team of two to twenty, roughly eight of the twelve features above will shape your working week, and the rest are insurance. Buy the routing, the CRM sync, the transcript archive, and the GDPR paperwork. Everything else is a conversation for year two, when you know what your queues actually look like.
Feature bloat is not a technology problem. It is a procurement one, and it costs you the margin you were hoping live chat would add.
