The Missing Leading Indicator

Why the handover of a won deal is the one point in the Bowtie where nobody measures what arrives.

Matthias Drebes · Founder, handedover · 1 October 2026 · Download the PDF

In his book Revenue Architecture (October 2023), Jacco van der Kooij described the Bowtie model for managing recurring revenue businesses (van der Kooij, 2023). Winning by Design has since published it as a proposed standard (Winning by Design, 2026). What is new about the model is its end-to-end view of the customer lifecycle. Instead of treating the close as the end of the sales cycle, the model extends into delivery (onboarding, adoption and expansion) and focuses on the whole customer lifetime value. The perspective shifts from the value and benefits of the software to the impact it has on the customer's business. If a customer's expectations are not met during the initial rollout and adoption, the deal was barely worth anything.

Figure 1. The Revenue Bowtie. Redrawn after Winning by Design (2026), Revenue Architecture: Data Architecture. The Bowtie, a proposed standard, v.1.0, pp. 6 and 44. Proportions follow the standard's specification (Annex A, p. 43).

What the Bowtie gives GTM and RevOps organisations

For GTM and RevOps leaders, the Bowtie does five things.

One system

Every stage is treated like a machine with an input and an output. This creates a standardised data structure that measures volume, conversion and time at each stage and makes data across marketing, sales and customer success comparable (Winning by Design, 2026, pp. 16 to 19).

Oversight across functions

Each measurement shows the performance of its stage, and together they show performance across functions. Leaders can identify and quantify the levers of improvement at each stage and their impact on overall performance (Winning by Design, 2026, pp. 26 to 27).

One language

Winning by Design pairs the model with SPICED (Situation, Pain, Impact, Critical Event and Decision), a qualification methodology that establishes a shared language and passes on what is learned across stages.

Benchmarks

The standard language allows benchmarking against a company's own historical data as well as against peers (Winning by Design, 2026, p. 27).

A basis for AI

A clean end-to-end dataset is the basis for AI initiatives and further automation in the GTM organisation (Winning by Design, 2026, p. 37).

The Bowtie's knot

At the knot, the deal changes hands. The customer signs, and everything the seller learned in months of selling has to reach people who were not in the room: what the customer expects, who really decides, what was conceded along the way. Winning by Design calls this stage Mutual Commit (Winning by Design, 2026, p. 7), and by its own account this is where the value begins: "most customer lifetime value is created after the initial sale" (Winning by Design, n.d.). The knot is the only entrance to that value.

The standard measures the knot with two numbers, and both have a clear job. The first is the price achieved: deals won times list price, minus discounts (Winning by Design, 2026, p. 21). It turns wins into committed revenue and shows how much was given away to close. The second is the time it takes to set the customer up in the company's own systems and processes (Winning by Design, 2026, p. 25): in practice, the contract in billing and the account in the CRM. It shows how quickly delivery can start, and speed matters here, because the longer onboarding takes, the more customers leave (Winning by Design, 2026, p. 25).

Both numbers describe the company's side of the deal: what it earned and how fast it is ready. Neither describes the customer's side: what they bought it for and what they expect to happen next. That is the side that decides whether they stay, and nothing at the knot measures whether it arrived.

An illustration. A deal closes at list price and is set up in two days. On both measures it is a perfect handover. But the customer signed expecting single sign-on in the second quarter, and nobody on the delivery side ever heard it. Both numbers stay green. The first sign of trouble is a disappointed customer, months later.

Both numbers stay green.

The first number that shows a failed handover is onboarding retention (Winning by Design, 2026, p. 21). By then it is already a result, blended with poor qualification and poor delivery. Even the standard credits better retention to better qualification (Winning by Design, 2026, pp. 26 to 27). Without a measure at the knot, nobody can tell the two apart.

The missing leading indicator

The model knows that early measures predict later ones: if onboarding takes days to weeks, onboarding churn rises (Winning by Design, 2026, p. 25). Yet at the one point where the right side begins, nothing looks forward. And a stage without a measure never shows up as the bottleneck.

Expectations are why this matters. Satisfaction is performance measured against the expectation formed beforehand (Oliver, 1980), and in B2B that expectation is set in the sale. What the seller set up is what the post-sales team will be judged against, whether they know it or not.

Documentation does not close the gap. SPICED defines what should be handed over, and a fully kept record is available. Available is not arrived. Some of what matters most is not in the record at all: the expectations set during the sale, the concessions made along the way, the politics, and the seller's judgement of who in the account really matters. Even complete content can fail to transfer. Szulanski found that knowledge transfer fails on the receiver's capacity to absorb it and on the relationship between the two sides, not on missing content (Szulanski, 1996). Customers feel the result as a break at the seam between teams (Tuli, Kohli and Bharadwaj, 2007).

Available is not arrived.

That makes handover quality a twofold leading indicator. In the short term it tells customer success and services whether the implementation will land. In the medium term it tells sales whether the account will renew and grow. Sales creates it and later lives from it, so sales has to own it.

The impact of the missing leading indicator

Many GTM executives I have spoken to over the last five months acknowledge that the handover is a weak spot, but they have not prioritised it until now, because no one quantifies it or knows how much it is hurting their organisation.

Other fields have measured it. In medicine, a structured handover programme cut medical errors by 23 per cent and preventable harm by 30 per cent; part of it is that the receiving doctor summarises what was handed over (Starmer et al., 2014). In B2B sales, no published study measures handover quality against an outcome. The nearest measured event is a salesperson leaving: the account's annual revenue falls by 13.2 to 17.6 per cent (Shi et al., 2017).

What changes with AI

The obvious objection is that AI now summarises every call and writes the handover document. But a summary takes the same input as the recording: what was said. Politics, the expectations that were set and the seller's judgement do not get into it. And an AI-generated handover document is available, but that says nothing about whether it arrived. More documents widen the gap between available and arrived.

At the same time, AI shifts work into the left half of the Bowtie, above all outreach and qualification. Winning customers becomes cheaper and faster, which makes the right side, and its entrance, more important. The only edge an incumbent has over anything newer is that it already knows this customer. That knowledge is created on the left and has to cross the knot.

What a revenue leader can do

A revenue organisation runs at least four kinds of handover every year: the won deal, a promotion, a departure and the territory recut. Few measure any of them.

The answer is one measure, taken on the day of the handover and owned by the organisation, on two levels.

Test it for one quarter against onboarding retention and time to first impact. Only data turns the measure into a leading indicator.

What does it cost you?

The executives I spoke to could not say what failed handovers cost them. That is the first number to change. Put a price on it in about two minutes, in your own figures and with our arithmetic, at handedover.ai/cost. No sign-up, and nothing you enter is stored.

References

handedover starts where this paper ends: with the first handover, run together with you, which shows for each account what was covered, mentioned or missing on the day it changed hands.

Matthias Drebes spent thirteen years in enterprise sales, seven of them at Celonis while it grew from 100 to 3,000 people. He has run both the giving and the receiving side of the handover of a won account. mat@handedover.ai