A two-week audit: 90 days of CRM data, 3 rep interviews, 4 call recordings, and a review of quotes and lost deals. Below is the real deliverable, with the company and people's names removed.
The four numbers above are read straight from the CRM. They are not calculated and not interpreted. Everything modeled sits in its own section and is tagged.
The left column is read from the CRM and the interviews. The right column is the target state, not a promise - every line has a measurement that will say after 90 days whether it happened.
Three steps that are the same in every project, whatever the findings are.
We go through the findings with the owner, confirm priorities, make the first week's decisions, and name who owns what.
One session · Week oneA team session and an owner session once a week - progress review, priority adjustment, clearing blockers.
Throughout the projectEvery action is marked not started, in progress or done. Updated every week. An Excel tracker the client can review.
Throughout the projectSystems that fix most of what process can affect at all. None of these actions depend on individual skill development.
The team has no defined weekly metric - how many quotes, how many meetings, how many qualified contacts. The last 90 days show 9 closed deals, but there is no data on the activity that produced them. So right now a conversion problem cannot be told apart from a volume problem, and no improvement is measurable.
Pick one main metric per rep per week (quotes sent, meetings held, or new qualified contacts). Post it where the whole team sees it. Every Monday - last week's number, this week's target. No extra reporting: one page, one number, one target.
More than 30 deals sit in a "thinking / maybe" status, worth about €1.7M in total. The split: 0 fresh (under 14 days), about a quarter at 14-30 days (~€400k), about a sixth at 30-60 days (~€300k), more than half older than 60 days (~€900k). The statuses have never been reviewed, so the CRM cannot say which deals are alive, which are dead and which were never really started.
All 30-plus deals get reviewed one by one with the rep who owns them. Three buckets: active, dead, never started. After that review the real open pipeline is known. This is also the only way to test the reactivation scenario modeled below - until then it is only an assumption.
The 90-day window holds about 50 quotes, of which 9 closed and about 40 were lost. That accounts for all of them, so the open count should be zero. At the same time the CRM shows 31 open deals worth about €1.7M. On top of that, the average open deal is about three times bigger than the average closed one, and the average closed deal is three times smaller. Two data sets with different deal profiles cannot go into the same calculation.
Before refining any euro model: find out which period and which stage each number comes from, and whether the large projects run through a separate process with a longer decision cycle. The likeliest explanation is that the large projects are a different type of deal, not "neglected" ones. Confirm this in the first 7 days, then recalculate the model ranges against the corrected data.
About 50 quotes, 9 closed, which is 17%. Of the lost deals, about 8 of 10 are tagged as "competitor was cheaper", with an average price gap of 15-20%. The company sets a margin floor a rep cannot go below, so fighting on the number is a closed road. Three interviews show three different reactions to price: discounts in about 30% of cases, arguing with certifications, and material optimization.
A conversation structure that moves the buyer from price to value: questions about the building's life cycle, about what a lower-quality solution costs later, about what the warranty is worth, about why the delivery schedule matters. This happens during the conversation, not before the estimate. The measurement is not "process rolled out", it is discount frequency, average margin and conversion against the 17% baseline.
"Competitor was cheaper" is the rep's note in the CRM, not a reason the buyer confirmed. A price note often hides the real reason: the spec, the timeline, trust, or weak qualification. So this finding is a signal, not a proven cause, and that is exactly why the loss interviews sit in this same phase.
From the 3 interviews: one follows up "whenever it happens" and pushes the task out 1-2 months if the buyer goes quiet. The second calls five times with no structure. The third contacts the existing base 1-2 times a year. There is no written sequence and no closing message. That is a fact about the process, not an interpretation.
A 7-touch sequence (1, 3, 7, 14, 30, 60, 90 days) with specific message templates, each touch with its own goal, from "did the estimate arrive" through to the close. CRM automation with reminders. Execution is measurable right away (how many quotes got the full sequence), the result only after 90 days.
In all 4 calls reviewed, the opening starts with a pitch of the company's offer. One of the four booked a meeting, and in that case the buyer took the initiative. The same pattern shows up in how the reps describe their own calls in the interviews.
An opening script with 3-4 questions before any pitch, plus a set way to ask for the meeting with a low-risk next step. In parallel, regular call reviews start so the sample gets big enough. The measurement: meeting booking rate before and after, on a comparable call volume.
Four calls are an indicative sample, not statistical proof. On that basis you cannot claim the whole team works this way, or that the opening is what drives conversion. The first step is a bigger sample: 15-20 calls a month scored on one shared sheet, before anyone draws conclusions from it.
Of the lost deals, about 8 of 10 are tagged as "competitor was cheaper", with no confirmation from the buyer. Without this review, price stays the catch-all answer that hides the real problem, and every other finding in this section stays untested.
For every lost deal, a short interview call to the buyer or an email with 3 questions about the real reason. The weekly pipeline review includes loss analysis. After 30 days there is real data that either confirms the price hypothesis or kills it. If it kills it, the value conversation drops in priority and the model gets recalculated.
Two items that deliberately carry no euro figure. Both lack the historical data a model would need, so they are delivered as a process with measurable execution, not as a revenue promise.
The company closes about 36 deals a year. There is no post-close process at all: across the 3 interviews, nobody mentions a referral ask or any structure. The rep with 2 years of experience contacts the existing base 1-2 times a year, but with no specific ask.
A 30-day post-close call script with 3 questions, plus a separate referral loop in the CRM. In the first 90 days the goal is not revenue, it is baseline data: how many calls were made, how many referrals came in, how many of those turned into meetings. After that, with real numbers, you can build a model that means something.
This finding deliberately carries no euro figure. Modeling it would need the company's historical referral conversion, customer satisfaction data and spare capacity. None of that exists. Public industry averages are not comparable, because they measure different things in different markets, so they are not used here.
Across the 3 interviews and the 4 call reviews, each rep shows a different gap: one skips discovery, one does not ask problem-finding questions, one never checks the decision-making process. The sample is small, so this is a starting point for individual work, not a diagnosis.
In week 1, a one-on-one session with each rep: we define the single question they will drill for 30 days. Each one gets 3 variations of that question, 15 min a day, a weekly note on Fridays. At the end of week 4, a 15 min check against 4 criteria (drilled 4 days out of 5, used it in 80%+ of calls, can describe one time it worked and one time it did not). If yes on all 4, they move to Level 2.
These are the only euro figures in the audit. Each one shows its formula, the assumption behind it, and the test that would kill it. All numbers are revenue, not profit. The ranges are wide on purpose - a narrow range would imply a precision that is not there.
~€400k × 25% + ~€300k × 15% + ~€900k × 8% = ~€217k gross, with a 50% discount for unverified CRM statuses = ~€110k
Assumes the deals are alive and reachable, and that the close rates by age band are close to true. None of the three rates comes from this company's own history, they are assumptions. The number is revenue, not profit: at roughly 30% margin, ~€110k means about €34k of gross profit.
If the 60+ day band delivers under 3% after the status review, the problem is not follow-up but qualification or price, and item A falls away entirely.
About 50 quotes a quarter × (target % − 17%) × 4 quarters × about €17.6k. At 19% = about €70k, at 22% = about €175k
Assumes quote volume and average deal size stay flat, and that the process changes move conversion at all. How much comes from the conversation structure, how much from follow-up and how much from qualification cannot be separated, so it is not separated. The second weak point: the calculation uses an average closed deal of about €17.6k, but the extra deals won could be systematically smaller or larger than the average, and the median was not calculated for lack of data.
If conversion on new quotes is not above the 17% baseline after 90 days, the item is zero. If margin has dropped, it is negative, even if conversion went up. If quote volume has fallen well below 50 a quarter, the calculation has to be redone from scratch.
No calculation. The company's historical referral conversion is unknown, and satisfaction data is not collected.
Any number here would be invented. Baseline data gets collected in the first 90 days, the model comes after that.
Nothing to kill, because nothing has been claimed.
The modeled numbers are not a forecast and not a guarantee. They are revenue, not profit, and the engagement fee is not subtracted from them. They show what would happen under stated assumptions, and every assumption is written next to the number. The ranges are wide because the data is incomplete, and they narrow only after the first 30 days, once the CRM is clean. If an assumption fails the test, the number gets recalculated, not defended.
Two weeks, fixed price. CRM data, team interviews, call reviews, and a model that states every assumption and the test that would kill it.