They owned every tool. They used none of them.

1. Populate the CRM from where the work already happened

People weren’t in the CRM. They were in their inbox, on WhatsApp, on SMS. So we stopped asking them to be somewhere else and brought the data to the system instead.

We pulled from the communication channels the team actually used - email, WhatsApp, SMS -and automatically populated the mother CRM from them. A portion of that activity was synced back out to the modules that needed it: marketing, customer satisfaction, customer success. The record filled itself as a by-product of people doing their jobs, not as a tax on top.

This is the unglamorous foundation. You cannot align a team around data nobody enters. So we made the entering disappear.

2. Find the few KPIs everyone would actually agree on

Adoption is not a training problem. People use a system when it measures something they care about. So before touching a workflow, we isolated a small number of KPIs that carried shared importance across every major stakeholder - the CEO, pre-sales, customer success - and got a definition each of them would sign.

That last part is the whole battle. A KPI everyone interprets differently is worse than no KPI. Once the definitions were agreed and adopted, those few numbers became the north stars. And here is what changed: the teams started interacting with the system, on purpose, because the system was now how you moved the number you were judged on.

The tools didn’t change. The reason to open them did.

3. Let the forecast lie, then make it impossible

The first thing that happened when sales had a KPI was that sales gamed it. Deals that were going nowhere stayed marked “in progress” because a fuller pipeline looked like better forecasting. Entirely predictable. Entirely human.

So we tightened the definitions until the gaming had nowhere to go. What counts as a commit. What a deal that is genuinely still alive looks like versus one that is just being kept warm on a slide. And, most importantly, a written definition of what “stalled” and “lost” actually mean - with the workflow and rules to enforce it.

The result looked, for a moment, like a catastrophe: a massive drop from “opportunities in progress” to “closed lost.” The pipeline appeared to collapse. It hadn’t. It had just stopped lying. For the first time, the number on the board was one the CEO could actually plan against.

4. Read the losses, then fix what caused them

A pile of honestly-labelled lost opportunities is not a failure. It is the best dataset a go-to-market team will ever get. So we analysed it - where, and why, deals were actually being lost.

Three things fell out of that. Communication practices that were quietly costing deals, which we addressed directly with internal workshops in a cross-functional room. Work that was pulling sales away from clients and into irrelevant admin, which we handed to automation. And a main-CRM view so cluttered that reps avoided it - which we simplified down to the few objects they actually needed to touch, made clean enough to use in seconds, on the go.

Adoption stopped being a mandate. The system was now faster than the workaround.

5. The parallel NPS project -and the 34%

Alongside the sales work, we ran an NPS programme with the customer success teams and their account managers - not as a satisfaction survey to file away, but as a reactivation engine.

It reactivated a meaningful set of dormant customers. That single campaign opened 34% new pipeline. It also did two things nobody expected: it surfaced exactly where the delivery teams had failed to produce the results that had been promised, and it opened honest conversations about optimising and updating their workflows and project-management programmes - problems that had been invisible while everyone avoided the system that would have shown them.