We built our own sales console instead of buying one

Hey there,

David here.

Everything we build starts from the same question. How does a client get better results than they are getting now.

That question built the delivery side of our console first. Then we turned it on selling, and it got harder to answer. We could tell a client what was landing at the top of their funnel. Telling them which part of their selling was losing the deals was guesswork.

So we built the screen that answers it, and we ran it on our own pipeline first. Testing it on ourselves is the only honest way to hand it to a client.

The full tour is in the video below. It runs fourteen minutes and shows every part of it working. What follows is the reasoning the video does not slow down for. It holds up on its own if you would rather read.

Watch: A Full Tour of the Sales Console We Run Kinetyca On

01What earns its own screen

One test comes before anything gets built, for us or for a client. A job earns its own screen when you would have to check it daily to run well. You already know you will not.

I cannot open the HeyReach dashboard every morning. Or EmailBison, or the CRM and the mailbox before a call. The information exists in all of them. It never arrives in time to change what I do that day.

So the console holds four decisions, each with a screen behind it.

• Conversion metrics decide which channel gets more budget
• The disco call prepper decides what I open a call with
• Follow ups decide who gets an answer today
• The sales call trainer decides what changes on the next call

Those decisions were never late because the data was missing. They were late because it sat in four places, and nobody opens four dashboards before lunch. That is where the speed comes from. A decision lands the same day instead of the following week.

02The counting mistake underneath most funnel reports

This next part is worth having even if you build nothing.

Most funnel reports count positions. They show where every deal sits right now. A deal that reached negotiation in March and closed in April stops counting as one that ever reached negotiation. The middle of the funnel looks thin because the successful deals moved out of it.

We count stage history instead. Every deal counts at every step it ever reached. The pass-through rate between two stages becomes the real one. It usually looks nothing like the report you had before.

Events then sit on top, with timestamps. A no-show, a rebook, a disqualification. When they happened tells you which lever to pull, and a position-based funnel cannot show you that at all.

Two things fall out of this on their own.

• Show-up rate and qualification rate move weeks before closed-deal data does. They are the early read on whether the message and the ICP are right.
• The channel with the most replies is often not the channel with the most revenue. Win rate and won value per channel are the only place that shows up.

Any CRM holding stage history can be read this way. Most of them make it genuinely hard to configure. That is why so few sales leaders I speak to can state their conversion numbers precisely. The number exists. It is being counted wrong.

03Where we stopped the automation on purpose

Two rules hold the rest together, and both are places we chose not to automate.

Nurturing before negotiation runs on its own. Guides, skill files, automation ideas, all on a sequence. Once a deal reaches negotiation that stops working, because the details get too individual. So the screen shows who has been waiting longest for an answer. It drafts the message inside the real thread. We adjust the tone or rewrite it, and then a person sends it.

The second rule is what makes AI call feedback usable. Every score has to cite a timestamped quote from the call, or the score does not count. Without that, you get feedback that sounds right and cannot be checked. With it, you get a line, a minute mark and a reason you can argue with. Gong publishes 43 percent as the talk ratio on winning calls. We hold our own scoring against published benchmarks like that.

Gong does this work very well. If the price fits what you need, use it. We wanted it shaped to how we sell, sitting next to everything else we run. So we built that piece ourselves.

One note on the build, because people ask. This reads the sending tools, the CRM, the calendar, the mailbox and our own database. The credentials never touch the model. That was a requirement before it was a feature.

We build versions of this for clients who want it shaped to how they sell. We pull their data from whichever tools hold it, and show it back the way they read best. That puts them ahead of a team without it.

One question back, and I would genuinely like the answer. Which decision in your sales week is late, only because the answer lives somewhere you cannot open daily?

Reply and tell me. I read them.

PS. If you want this shaped around your own pipeline, grab a time here.