How Jobilla went from tools that captured everything to a co-worker that does the work

Jobilla is a recruitment marketing company. It helps employers reach and convert candidates who are not looking for a job, and it sells into employers across Europe. The sales organisation runs on HubSpot and sells in trios: a BDR who opens the door, an Account Executive who sells, and an Account Manager who takes the opportunity on after the close and grows it.

Ask Tommi Siro, Co-Founder and CGO, what Jobilla is really trying to achieve, and he does not start with tools. He starts with how the money arrives.

A large employer does not buy recruitment marketing centrally. One unit does. A single site, a single country, a single business line inside a company with forty more of them, each with its own budget and its own hiring problem. Land that one, prove it, and the account is worth many times what the first deal was. One account Jobilla is working now is a charitable organisation with roughly three hundred local branches, each hiring on its own budget.

That is the symptom. The disease is that a motion like this asks people to remember an enormous amount: which units have been approached, what each conversation established, which of the three people on the account owes the next move.

Put numbers on it. A Jobilla seller carries around twenty open deals at any moment, each one a unit inside a larger company with its own stakeholders and its own overdue next move. Across the team that is over a thousand meetings a month, more than fifty conversations every working day, plus the emails and notes attached to every one of those deals.

Being properly prepared for a meeting means re-reading the last conversation on that account, the thread since, and whatever happened on the neighbouring units of the same customer. Fifteen minutes of reading before a thirty-minute call, several times a day. So it does not get done, and the rep walks in with whatever they can recall.

For a manager it is worse, because the most valuable question, why one seller keeps winning where another does not, needs the most reading of all. Tommi is the person at Jobilla who most wanted to answer it, and he had three years of recordings to answer it with. He got through three meetings a month. The other thousand went unwatched, so coaching ran on impression, on whichever deals came up in a pipeline review, and on whoever spoke loudest about their own quarter.

None of this is a discipline problem. There is no version of a human being who keeps twenty deals current by remembering to. Three things, as Tommi puts it, that Jobilla has struggled with for as long as it has existed: the ball stays in the air, the data stays clean, tracking and execution stay systematic. All three depend on somebody remembering. There is exactly one other way to get them, and Tommi has a name for it.

The axe is what most sales organisations quietly run on, and there is a paradox underneath it. The better a seller does, the more of this work they generate: every deal opened is another record to keep current, another set of follow-ups to log. Sell more and you earn more of the task you least want, and in plenty of companies letting it slip is a serious offence. So the people best at the job carry the heaviest version of the thing they like least, the admin work.

This was not for want of trying. They ran a meeting recorder, and it did what meeting recorders do: it captured the call, and then it waited for somebody to come back to it. They built automations into the CRM and calendar reminders that were supposed to keep the records current.

None of which was for want of a good CRM. Jobilla's is well built and carefully looked after. But a CRM is a static database by nature. It holds what somebody has already put into it, which means every field in it stands for work a seller did, rather than work a seller was spared.

Every one of those tools captured something. Not one of them decided anything. Each made the same silent assumption, that somebody would go back and read it, and the arithmetic above is why nobody ever could.

Two numbers explain most of what Ollie does at Jobilla, and the gap between them is the point.

Every month, Ollie records, transcribes, analyses and acts on over a thousand meetings for Jobilla, using its own recorder as the way in, and reads about six hundred emails into the same picture. Every tool Jobilla had already tried did some version of that.

What none of them did is the second number. In the same month, Ollie completes over two thousand pieces of finished work for the team: roughly eighty per seller, four every working day. Finished meaning the email is written, the CRM update is drafted, the deal is created. Work a rep reads and approves in Slack, rather than work a rep still has to do.

What that work actually is

Everything below arrives complete. None of it is a prompt to go and do something.

A brief before the day starts. Prep before every customer call, built from every previous interaction on the account and fresh research on the people in it. After the call, the write-up, the next steps, and the CRM brought up to date from what was actually said. Where a deal should exist and does not, the deal itself, drafted from the conversation and the people in the room, with contacts attached and a close date proposed, for the rep to approve in Slack.

Then the part that matters most and gets talked about least. Every promise a rep made in a meeting or an email is tracked, pulled out of their own conversations, and closed off when an email, meeting transcript or calendar show it has been done. A rep does not have to remember what they committed to on a call three weeks ago, because Ollie was in the meeting and wrote it down.

A flag when an inbound email matters more than it looks. A nudge when an account has gone quiet, with the re-engagement already written. A value summary or a business case when a champion needs help to sell internally. And for the people running the team, a weekly read on what moved, what stalled and why, and an analysis of every deal as it closes.

Tommi, on what that feels like from the inside:

Watching the Friday check get configured, Jobilla's sales leader named exactly what it replaced:

A reminder had existed because the thing kept not happening.

The work above arrives whether or not anyone asks. What tells you the team actually trusts Ollie is what they go to Ollie for.

It starts with the pipeline. The first question a sales leader asks is the one that used to take a week of reports to answer: where are we, and how does the month end? Ollie reads the whole pipeline and answers in one pass.

• What is open and what it is worth.
• How much of it is at risk, and why.
• Which deals are due this month, and which of those are still likely.
• A plain verdict on whether the month looks soft or solid.
• And then the part no report ever gave anyone: which deals can still be saved before the month closes.

From there the questions get sharper, and they are the questions a leader actually has.

None of that is a dashboard. Each is an answer, built from what was actually said, to a question nobody had time to research.

Two of those questions, in detail

Where is the team stuck? Tommi asked what was blocking one of the regional sales teams. The answer did not flatter him. It told him the pipeline was in bad shape, named the share of deals at risk, and counted the ones that had gone quiet. More useful than the count was the pattern. The top blocker was not the market or a competitor. It was a habit: reps sending a proposal and then going passive, deals dying in silence, across the whole team. Underneath that, deals stuck with contacts who could not sign, accounts frozen for the summer with nothing prepared for their return, and a few technical blockers that needed coordination rather than waiting.

It also named a benchmark, one rep whose deals were in far better shape than the rest, which is the evidence that the pattern is a habit and not the market. And it said where to start. That is a coaching plan, not a report.

Where is the rest of this account? The hardest question in a land-and-expand motion is the one nobody has time to answer. Take one global manufacturer Jobilla sells to. Head office buys a package and the local units draw on it, so the account runs across a dozen deals in eight countries. Tommi asked Ollie how it was doing worldwide. The first answer came from the deal records, and Tommi pushed back: on an account like this the buying is central and the work is local, so the deal records show what was bought, not what is happening. Ollie went to the meetings instead, thirty-two of them across seven of those countries, and sorted them by which ones needed Tommi.

Two units needed attention that day. In one, the contact who had promised an update by September had gone silent with no follow-up sent. In the other, hiring had stalled. Two more were worth watching, including an expansion parked on somebody's holiday since June. Three were running well. And underneath it all, a warning: the customer had begun a global vendor evaluation, and the time to get close to the champion was now.

That picture lives in thirty-two conversations held in seven countries over months. Assembling it by hand was never going to happen. It took one question.

Ask Tommi what he can see now that he could not before, and he corrects the question:

The information and the recordings always existed. What did not exist was any way for one person to get through them.

Ollie arrives where the team already is. The work lands in Slack, drafted, ready to use. HubSpot stays the system of record throughout; what changes is who has to keep it current.

And the work lands on the rep's side of the table. A co-worker that sits in every meeting raises an obvious worry, and it has a name on every sales team: big brother. What defuses it at Jobilla is not a policy. It is that every day, the thing in the meeting hands the rep the deal already created, the update already drafted, the follow-up already written, and the promises already tracked. Hours of their own work, done before they asked. A rep does not experience that as being watched. They experience it as being helped, and the recording becomes the price of the help rather than the point of it.

Tommi's own position closes the rest of the gap:

There is a version of this technology built for watching people. Ollie is not that, and the clearest proof is a behaviour rather than an argument: Jobilla's reps ask Ollie to critique their own calls. Nobody volunteers to be surveilled. People do volunteer to be helped.

The pipeline they can see has more than doubled

Since Ollie took over creating and updating deals, every qualifying deal is created on the day of the meeting that qualifies it, rather than once business already looked likely. Jobilla now creates 60% more deals every month, and the total value in the pipeline has risen by 108%.

The gap between those two numbers is the finding. Value grew nearly twice as fast as volume, which means the average deal now being tracked is around a third larger than before. Whatever had been going unrecorded, it was not the small change. Real business was invisible until it was nearly closed, and a forecast built on the visible half was never going to be right.

None of this is new demand. It is the same business, finally visible while it is still in play. That is the thing Tommi had wanted fixed for ten years.

Reps keep all the promises they make

Nobody got there by trying harder; the remembering simply stopped being a person's job.

The holidays show what that means. Reps remember their biggest deals, and nobody needs reminding about the account that decides their quarter. What slips is the smaller opportunity where somebody said "let's pick this up after the break." That promise falls due weeks later, to a person who has been away and has had twenty other deals in front of them since they got back. Those follow-ups used to quietly not happen, and nobody could point to the loss because nothing visibly broke. This time the commitment was captured the moment it was made, and it came back when it was due.

Every seller works with Ollie, every day

Not most of them, and not weekly. All 25 work with Ollie daily, in every market Jobilla sells into. For a team that had three earlier attempts fail, that is the number that should be hardest to believe.

Tommi's reaction on the day deal creation went live, mid-sentence and unprompted, after ten years of the problem: