Kestrel Analytics: moving upmarket to fix a retention problem

Is retention a segment problem?

The evidence does not support the assumption that enterprise will fix retention. NRR of 96% means existing customers are contracting or leaving faster than others expand. 14% logo churn in SMB usually signals a mismatch between price, onboarding and realised value. If the product does not embed in small-customer workflows, it will not embed in enterprise pilots either - enterprise buyers are less forgiving, not more.

Ansoff framing

Selling the existing product to larger customers is market development, moderate risk. But the product is not in fact existing: SSO, audit logs, data residency and a 99.95% SLA are new features. That makes the move effectively diversification - new customers plus new product - the highest-risk quadrant, and it should be resourced and governed accordingly.

Unit economics

61% gross margin and 22-month CAC payback is thin. Enterprise typically brings a 9–12 month sales cycle and higher CAC, offset by larger contract value and real expansion potential. That payoff depends on a land-and-expand motion Kestrel has not yet demonstrated.

Cash and capacity

£3.5m against £6m cash consumes over half the balance sheet, with no enterprise revenue until year two. A slip starves the SMB business - the current cash engine - at the same time.

Judgement

Do not pivot wholesale. Fix retention first (onboarding, activation, success management), then pursue two or three lighthouse enterprise accounts, buying rather than building compliance features where possible. Set a gate: £2m of enterprise ARR within 12 months or the initiative stops.

Measures that fit the strategy

The current system measures acquisition, which is precisely what produced the problem. Add:

• NRR and GRR by segment, monthly - the primary health measure
• Time to first value (days from contract to first live workflow) and activation rate
• Expansion ARR, and expansion per account manager
• CAC payback by segment and gross margin after hosting and support
• SLA attainment and ticket deflection - required to make the enterprise promise credible
• Pipeline coverage by deal-size band and cycle length, since enterprise conversion assumptions are untested

Reward design

Pay sales on total contract value including year-two expansion, with a separate expansion quota owned by a customer success team. Cap new-logo accelerators so reps cannot buy logos with discounts. Add a quality element: commission reduced if an account churns within twelve months. Move part of the CTO's objectives from delivery dates to feature adoption.

Behavioural risks

Any comp change invites gaming. Reps may sandbag pipeline, push multi-year deals to inflate TCV, or abandon small accounts that no longer earn commission. Expansion targets can encourage over-selling features that later increase churn. Mitigate with GRR as a counterbalancing measure and a monthly customer-health review chaired by the CEO. Define every metric once, freeze definitions for twelve months and audit a sample quarterly - metric churn mid-transformation is worse than having no new metric at all.

Diagnosing the change

Using a force-field view, the driving forces are worsening retention, board pressure and two live enterprise prospects. The restraining forces are more numerous: a £6m cash balance, an engineering organisation optimised for rapid SMB releases, a sales force paid on logos, no security-cleared staff, and a proposition with no compliance features.

Using Kotter-style sequencing, the gap is at step one: there is no shared sense of urgency beyond the boardroom, and the CEO's 30%-of-ARR target is a destination, not a case for change. Staff will hear 'we are abandoning SMB' unless the retention diagnosis is communicated honestly.

Sequencing recommendation

1. Months 0–3: fix the leak first. Establish NRR as the company metric, appoint a customer success lead, redesign onboarding. Build credibility before asking for £3.5m.
2. Months 2–6: build a guiding coalition spanning product, security, sales and finance, with a single accountable executive for the enterprise programme.
3. Months 3–9: lighthouse accounts. Pursue the insurer and the NHS trust as design partners with a paid pilot, using partner technology for SSO and audit logs rather than building everything.
4. Months 6–12: change the reward system once the operating model exists - not before, or reps will game an undefined target.
5. Month 12: gate review against enterprise ARR, NRR and cash.

The main risk

Running two operating models on one P&L with £6m of cash. If the board insists on the CEO's two-year target, it must also accept a slower SMB roadmap and a hiring freeze elsewhere - an explicit trade-off rather than an assumed one.