INCGPT: from market interest to a working pipeline

What we found

INCGPT is an AI startup. The product worked and the market was interested, which is the best position a young company can be in and also the one that hides the most. Leads came in through the website, content, the network and LinkedIn. Four channels, none of them wired into a process.

Everything landed in a single list. A visitor who read an article sat next to a buyer who had asked for a demo, with the same status and the same weight. With no distinction between a marketing qualified lead and a sales qualified one, the list could not be sorted by intent. It could only be read from the top until the day ran out.

Forecasting followed from that. It ran on feeling, and a feeling cannot be broken down into stages, owners and dates. Strong product, real demand, missing foundation.

Why a flat list breaks the forecast

A forecast is arithmetic over records that all mean the same thing. Once the set holds two populations, an interested reader and a buyer in a process, counting them together overstates the pipeline and counting only the obvious buyers understates it. Both are wrong in a direction nobody can predict.

That is the practical reason the MQL and SQL distinction exists. It is not vocabulary. It is the line between the people marketing is still working on and the people sales has accepted, and it is what makes a forecast possible at all.

The foundation we built

At INCGPT we build the RevOps foundation first, from lead registration through to a qualified pipeline, with automation as the mechanism rather than as decoration on top.

First the states, where we design the lifecycle so a record holds one status at a time and one clear reason to move: lead, then MQL, then SQL, then demo, then proposal, then customer. Each state carries a written definition of what has to be true before a record may sit there. The definitions are the deliverable. The configuration follows.

Then the account, where we configure the deal stages to match those definitions exactly, so the board and the lifecycle tell one story instead of two. Where a transition can be decided by a rule, the rule does it. Where it needs judgement, a person does it and the account records who.

Connecting the channels

Next the intake, where we connect every channel that produced a lead into the same record set: website forms, email, and the outbound tooling. A channel that keeps its own list cannot be compared with any other, and four private lists produce four versions of the quarter.

With intake in one place, which channel produces leads that reach SQL becomes an ordinary query. That layer is on the lead inbox page, and the same data is what makes attribution more than a guess.

Scoring, outbound and the pipeline

INCGPT runs AI driven outreach with automatic lead scoring, which is useful as long as the score lands somewhere it can be acted on. So the scoring is connected to the pipeline instead of sitting in a separate tool with its own view of the world.

That connection keeps outbound honest. A score raises a record's priority, and it promotes nothing by itself, because a model that both generates the contact and grades the contact will eventually grade its own work generously. The volume side is covered in outbound at scale.

What this case does not prove

We publish no numbers for this project, and the absence is deliberate. Percentages are easy to write and hard to attribute honestly at a company that kept growing while the work ran. What we state is what we built, and in what order.

The case is also a poor argument for anyone whose problem is demand. A RevOps foundation organises demand, it does not create it, and a company without leads has built a filing system for an empty drawer.

The last caveat is the familiar one. A structure only holds if the team keeps using it, which is habit rather than configuration, and it is measured rather than assumed. What that measurement looks like is on the adoption rate page. Across more than 150 implementations as a Pipedrive Platinum Partner, the accounts that lasted were never the cleverest builds.

Sales Surge built this foundation. We do the implementation and we build the software that sits around it, which is why the channels in this account write into one place instead of four. If your leads arrive through five doors and none of them are counted the same way, that is where we would start. The rest is on sales-surge.io.

Questions