Reference
Worked Example: Sablecrest Software
Part of the Strategic Choice skill · loaded on demand from SKILL.md
Contents
- Client situation and fact base
- Stage 1: the framed choice
- Stage 2: four possibilities
- Stage 3: conditions for each option
- Stages 4 to 6: barrier conditions, tests, results
- Stage 7: the chosen cascade
- Stage 8: activity-system map, fit tests, trade-offs, signposts
- Stage 9: the choice record and hand-offs
1. Client situation and fact base
Sablecrest Software is a privately held provider of cloud quality-management software (QMS): six modules covering document control, training records, corrective and preventive action (CAPA), audit management, supplier quality and change control. It is headquartered in Raleigh, North Carolina, with an engineering hub in Dublin, employs 540 people, and has been majority-owned since 2022 by a growth-equity firm, Marchbank Capital, which expects an exit in fiscal 2028 or 2029. Fiscal 2025 closed with annual recurring revenue (ARR) of 96.0 million dollars, recognised revenue of 91.0 million, ARR growth of 10 percent, net revenue retention (NRR) of 103.6 percent, gross margin of 74.3 percent and run-rate adjusted EBITDA of 7.1 million (7.4 percent), a Rule of 40 score of 17. The market for cloud QMS is growing at about 18 percent; the two enterprise platforms that dominate the top of it grow at 25 percent or more, and a wave of low-cost horizontal tools with free tiers takes small accounts from below. The board asked the CEO, Daniel Okafor, for a three-year plan. His first draft proposed "growth in every segment": a move upmarket into enterprise accounts, a free product-led tier for small companies, entry into three new verticals (food and beverage, aerospace, automotive) and AI copilots in every module. The board chair, Catherine Voss, an operating partner at Marchbank, asked what the plan excluded. It excluded nothing, which is why the engagement started.
Fact base assembled by the upstream skills (value-driver-modeling for the segment economics, market-competitive-intelligence for the segment sizing and competitor profiles, primary-research-synthesis for 30 quality-leader interviews and 30 churned-account interviews):
| Segment | Logos | ARR (M) | ARR per logo (k) | Gross margin | GRR | NRR | S&M (% ARR) | Contribution | CAC payback | Trend |
|---|---|---|---|---|---|---|---|---|---|---|
| Regulated life-sciences mid-market (medical device, pharma, biologics, contract manufacturers; 200 to 2,000 employees; sales-led) | 300 | 28.8 | 96.0 | 80% | 96% | 118% | 26% | 54% (15.6M) | 18 months | Growing 22% |
| Horizontal mid-market (industrial, food, automotive, services; sales-led) | 780 | 51.2 | 65.6 | 75% | 88% | 101% | 34% | 41% (21.0M) | 28 months | Growing 9%; losing about 30 logos a year to the enterprise platforms and low-cost tools |
| SMB and self-serve (under 200 employees, mixed industries; inside sales plus human-supported onboarding) | 2,100 | 16.0 | 7.6 | 62% | 77% | 86% | 48% | 14% (2.2M) | 47 months | Shrinking 4%; gross margin down 5 points in two years on support cost |
| Total | 3,180 | 96.0 | 30.2 | 74.3% | 88.6% | 103.6% | 33.9% | 40.4% (38.8M) | ARR growth 10% (from 87.2M) |
Definitions used throughout. Contribution is gross margin less sales and marketing allocated to the segment; research and development (22 percent of ARR, 21.1 million) and general and administrative cost (11 percent, 10.6 million) are unallocated, which gives run-rate EBITDA of 7.1 million. CAC payback is sales and marketing spend divided by gross new ARR (new-logo plus expansion) times gross margin, expressed in months. GRR is gross revenue retention (ARR retained from the year's opening cohort before expansion). Percentages in the P&L are on year-end ARR run-rate, which is how the board tracks the business.
Market facts: North America and Europe contain about 6,800 life-sciences manufacturers with 50 to 5,000 employees. Of these, about 2,900 have 200 to 2,000 employees and annual quality-software spend of 120 to 300 thousand dollars (average 200 thousand), a segment pool of about 580 million of ARR. About 2,450 of them (84.5 percent) run one of the three ERP systems for which Sablecrest has certified connectors, a reachable pool of 490 million. About 30 percent of the segment (roughly 870 companies) is on a modern cloud QMS with anyone: the two enterprise platforms hold about 400, Sablecrest 300, others 170. The enterprise platforms' sales teams concentrate on companies above 2,000 employees and deals above 250 thousand of annual contract value, and a typical enterprise-platform implementation costs a customer 350 thousand or more before validation. Sablecrest's vertical accounts average 2.4 of the six modules; the 40 accounts (13.3 percent) on the full bundle average 5.1 modules and 210 thousand of ARR. Interviews found that 23 of 30 quality leaders rank audit readiness and validation burden above licence price, and that validating a QMS release against the electronic-records rules they operate under (FDA 21 CFR Part 11 in the United States, EU GMP Annex 11 in Europe, with ISO 13485 governing device quality systems) takes their own teams 8 to 12 weeks per release. Of 30 churned SMB accounts interviewed, 13 (43.3 percent) had gone back to spreadsheets and shared drives or to a free tier of a low-cost tool.
2. Stage 1: the framed choice
"Should Sablecrest remain a horizontal quality platform selling to any company with a quality function, or become the quality system of record for one regulated vertical and give up the segments that choice excludes? If it specialises, on which basis?"
Okafor agreed that choosing one option forecloses the others: a vertical system of record cannot keep a horizontal new-logo sales team, a human-supported small-business tier, three new verticals and an "every segment" posture without straddling. The status quo was written up as a deliberate option (A) with its own trajectory.
3. Stage 2: four possibilities
Each option was written as a happy story by its advocate, without evaluation, and each names what it stops doing.
A. The horizontal platform done well (status quo). Sablecrest stays a horizontal QMS for any mid-sized company, takes 5 million of cost out of SMB support by automating onboarding, and invests 4 million in AI copilots across the modules to match the enterprise platforms' feature announcements. It wins on breadth and price against the platforms and on depth against the low-cost tools. It stops doing: nothing structurally; it reduces headcount by 35 in support and inside sales. Trajectory if conditions hold: ARR of 127.7 million by fiscal 2028 (growth of 12, 10 and 8 percent as horizontal churn continues and SMB shrinks), run-rate EBITDA of 7.0 million (5.5 percent). Advocate: Helen Strand, CFO, on the grounds that it is fundable and reversible.
B. Quality system of record for regulated life-sciences manufacturers (needs-based). Sablecrest becomes the QMS of record for medical-device, pharma, biologics and contract manufacturers with 200 to 2,000 employees that run one of the three certified ERPs: a pre-validated product that ships a maintained validation package with every release, a fixed 16-week implementation by GxP-certified consultants, certified connectors to ERP, MES and LIMS systems, sold as a six-module bundle on three-year terms. It wins because a company of this size cannot absorb an enterprise platform's implementation and validation cost and cannot get validation packages from a low-cost tool. It stops doing: new-logo sales to non-regulated horizontal accounts (3.7 million of new ARR a year), the human-supported SMB tier (16.0 million of ARR moves to a no-touch plan), custom development for companies over 2,000 employees, and module pricing below the bundle. Advocate: Okafor.
C. Best-of-breed supplier-quality point solution (variety-based). Sablecrest narrows to supplier quality management, the module with its highest satisfaction scores (14 percent of ARR, 13.4 million, at 84 percent gross margin), and sells it worldwide as a standalone product with product-led onboarding, integrating with the enterprise platforms rather than competing with them. It wins on depth in a module the platforms treat as an afterthought. It stops doing: the other five modules (86 percent of ARR) over three years. Advocate: Anika Sorensen, chief product officer.
D. Product-led long tail (access-based). Sablecrest serves regulated companies under 200 employees (device start-ups, contract laboratories, small pharma) with a self-serve product at 12,000 dollars a year, sold through marketplace listings and a channel of independent quality consultants, where the enterprise platforms cannot sell economically. It stops doing: sales-led mid-market selling and the field sales team. Advocate: Jonah Feld, VP growth.
Distinctness check: no two options could be pursued together without straddling. B and C conflict on product breadth and buyer; B and D conflict on segment and sales motion (validated sales-led against self-serve); A conflicts with all three on what is exited.
4. Stage 3: conditions for each option
Conditions were written by the whole executive team, with the sceptic for each option (the person who would vote against it) asked, "What would have to be true for you to support this?" Confidence was rated privately by each of six executives, then compared. Full sheet for option B; barrier conditions only for the others.
Option B conditions. Sceptic: Marcus Bell, chief revenue officer, who built the horizontal sales team and runs implementation, and who doubted the module-consolidation and consultant-hiring conditions.
| ID | Heading | Condition | Team confidence |
|---|---|---|---|
| B1 | Industry | The target segment (life-sciences manufacturers with 200 to 2,000 employees on a certified ERP) contains at least 2,000 companies and 400 million of annual quality-software spend | High (sized at 2,450 companies, 490 million) |
| B2 | Industry | Segment quality-software spend grows at least 12 percent a year through 2028 | Medium |
| B3 | Industry | Cloud-QMS penetration in the segment is below 40 percent and the enterprise platforms' sales model targets companies above 2,000 employees and deals above 250 thousand | Medium-high (measured 30 percent) |
| B4 | Customer value | Quality leaders in the segment will consolidate at least four of six modules on Sablecrest in return for a maintained validation package and certified connectors, up from 2.4 today, taking ARR per account from 96 thousand to at least 180 thousand | Low (Bell, Strand); medium (others) |
| B5 | Customer value | Quality leaders will sign three-year bundle agreements with a validation subscription priced 12 to 15 percent above module list, rather than insisting on module-by-module pricing | Medium-low |
| B6 | Customer value | Quality leaders rank audit readiness and validation burden above licence price when choosing a QMS | Medium (interviews: 23 of 30) |
| B7 | Relative position | Sablecrest can serve the vertical at 79 percent gross margin or better and a CAC payback of 20 months or less at 560 accounts | Medium |
| B8 | Relative position | Sablecrest can recruit or retrain and certify 48 GxP-qualified implementation consultants in 36 months at a loaded cost of 140,000 dollars or less each | Medium-low (Bell: low) |
| B9 | Relative position | At least 80 percent of target companies run an ERP for which Sablecrest already has a certified connector | High (measured: 84.5 percent) |
| B10 | Relative position | Moving SMB to a no-touch tier and stopping new-logo sales to non-regulated horizontal accounts retains at least 50 percent of SMB ARR and 70 percent of horizontal ARR by fiscal 2028, and cuts those segments' cost to serve by at least 40 percent | Medium |
| B11 | Competitor reaction | The enterprise platforms will not launch a pre-validated mid-market bundle below 100 thousand of annual contract value within three years, or if they do, cannot implement it in under 20 weeks without a new services model | Medium |
| B12 | Competitor reaction | The low-cost horizontal tools will not offer maintained validation packages or certified ERP connectors within three years | Medium-high |
Option A barrier conditions. A1: SMB gross retention stabilises at 80 percent or above once automated onboarding and the AI copilots are live (low confidence: it has fallen 6 points in two years). A2: neither enterprise platform launches a mid-market tier below 100 thousand within three years (medium-low: one has announced a "growth edition" for the next fiscal year).
Option C barrier conditions. C1: at least 35 percent of supplier-quality buyers at companies on another QMS platform would buy a standalone supplier-quality module from a second vendor (low). C2: the one module can carry the fixed cost of exiting the other five (medium-low: 13.4 million of ARR against 31.7 million of current research and development and general and administrative cost).
Option D barrier conditions. D1: regulated companies under 200 employees in North America and Europe represent at least 250 million of addressable ARR at a 12,000 dollar price point, enough to move a 96 million company (low; nobody had sized it). D2: self-serve onboarding without human implementation reaches a CAC payback under 24 months (medium-low).
5. Stages 4 to 6: barrier conditions, tests, results
Tests were designed by the sceptic for each option, pass marks set in advance, and run in order of lowest confidence first. Testing stopped for an option once a barrier failed.
| Option | Condition | Test (designed by) | Pass mark set in advance | Result | Verdict |
|---|---|---|---|---|---|
| D | D1 | Desk sizing from device-registration databases and company registers (Feld) | At least 250 million of addressable ARR | 9,400 companies at 12,000 dollars: 112.8 million | Fail; option eliminated in week 2 at a cost of three analyst days |
| C | C1 | 40 structured interviews with quality leaders at companies on a rival platform (Sorensen) | At least 35 percent would buy a standalone module | 8 of 40 (20 percent); 27 cited a single validated system of record as a requirement | Fail; option eliminated in week 5 |
| A | A1 | Cohort analysis of SMB retention and churn destination; 30 churned-account interviews (Strand) | GRR flat or rising in the latest two quarters | Latest two quarters annualise to 75 percent GRR, down a further 2 points; 13 of 30 churned accounts went to spreadsheets and shared drives or a free tier | Fail; status quo shown to be a managed decline, not a stable option |
| B | B4 | 90-day pilot: validated bundle offered to 10 vertical accounts at renewal, each on two or three modules (Bell) | At least 6 of 10 expand to four or more modules; average ARR per account at least 180 thousand within six months | 7 signed; 3 declined (2 had a corporate mandate for a parent group's platform, 1 wanted module-by-module pricing); average 4.3 modules and 192 thousand of ARR at month six | Pass |
| B | B8 | Post consultant roles in Raleigh and Dublin for 60 days; assess implementation staff for GxP certification (Bell with HR) | At least 12 qualified external candidates per hub at 140,000 or less; at least 15 existing consultants assessable as suitable | 11 and 8 candidates; 16 of 27 existing consultants suitable | Conditional pass: external hiring alone is insufficient; retraining closes the gap, leaving 32 external hires over 36 months |
| B | B5 | Bundle-pricing acceptance within the pilot (Strand) | At least 60 percent of signers accept the three-year bundle with validation subscription | 5 of 7 accepted (71 percent); 2 took module pricing at an 18 percent higher per-module rate | Pass |
| B | B7 | Segment model at 560 accounts, built in value-driver-modeling (Strand) | Gross margin at or above 79 percent; CAC payback 20 months or less | 79 percent; 17 months, driven by sales and marketing at 28 percent of ARR and 168 thousand of ARR per account | Pass |
| B | B11 | Competitor-response analysis handed to war-gaming (Bell) | No credible path for an enterprise platform to launch a validated mid-market bundle with a sub-20-week implementation within three years | Most likely response is discounting on Sablecrest's top 20 vertical accounts; a validated mid-market edition needs a separate services model costing about 6.0 million to build and about 2.4 million of new ARR a year at 50 percent contribution for a five-year payback, so a launch before 2028 is unlikely | Conditional pass; signpost set |
| B | B10 | SMB cohort by support-ticket volume and login frequency; horizontal run-off model at 90 percent NRR with no new logos (Strand) | At least 50 percent of SMB ARR and 70 percent of horizontal ARR retainable by fiscal 2028 | 56 percent and 74 percent estimated | Conditional pass; signpost set |
No pass mark was changed after results arrived. Bell, who designed the B4 and B8 tests, stated at the results review that the pilot changed his view on module consolidation and that the retraining route made the consultant condition acceptable.
6. Stage 7: the chosen cascade
Winning aspiration. To be the quality system of record for mid-market life-sciences manufacturers in North America and Europe: number one by count of customers among companies with 200 to 2,000 employees by the end of fiscal 2028 (560, from 300 today, against the enterprise platforms' combined 400), with NRR above 115 percent, run-rate EBITDA margin above 18 percent and a Rule of 40 score above 40. Excluded: being the largest QMS vendor by ARR or by logo count across all industries, which the enterprise platforms will remain.
Where to play. Customers: medical-device, pharma, biologics and contract manufacturers with 200 to 2,000 employees that run one of the three ERP systems with certified Sablecrest connectors (about 2,450 companies). Product: all six modules sold as one validated bundle. Geography: North America and Europe. Channel: direct, sales-led, with Sablecrest's own implementation, plus referrals from the three ERP vendors' partner programmes. Excluded: new logos in non-regulated horizontal segments; companies under 200 employees (existing SMB accounts move to a no-touch plan with a 6,000 dollar minimum); companies above 2,000 employees or any deal requiring custom development; the three proposed new verticals; any geography where none of the three ERPs is established.
How to win. Differentiation: the quality leader chooses Sablecrest over an enterprise platform or a low-cost tool because Sablecrest ships a maintained validation package with every release (cutting the customer's own validation effort from 8 to 12 weeks to under 2), implements in 16 weeks with GxP-certified consultants, and connects to the ERP the plant already runs, at a bundle price below what an enterprise platform charges for implementation alone. The advantage is sustained by the combination of release-level validation discipline and connector certification, which the enterprise platforms' mid-market economics do not justify and the low-cost tools cannot staff. Positioning base: needs-based, in Porter's terms.
Capabilities (must be distinctive). 1. Validation-ready release engineering: every release ships with installation and operational qualification documentation, a traceability matrix, and audit-trail and electronic-signature controls that meet the electronic-records rules (exists for the 40 bundle accounts as a manual effort; must be built into the release train, 4.2 million). 2. GxP-certified implementation and support: 16-week fixed-scope implementation, consultants certified and recertified annually (exists for 16 consultants; needs 48). 3. Certified integration layer: connectors to the three ERPs, plus MES and LIMS (ERP connectors exist; MES and LIMS to be built, 1.8 million). 4. Regulatory intelligence and benchmarks: a regulatory-change feed mapped to workflows and anonymised audit-finding benchmarks across the customer base (to be built; depends on the base growing). 5. Bundle-based account management: three-year agreements, module-adoption tracking, quarterly reviews on audit outcomes (partly exists; needs a playbook).
Management systems. Monthly account P&L for every vertical account, including implementation and support cost. Module-adoption tracking by account against the four-of-six condition. Validation-package sign-off as a release gate: no release ships without it. GxP certification curriculum with a 10-week course and annual recertification. Bundle playbook with approved price bands for the validation subscription. Sales and customer-success compensation on net retention and module adoption, not on new-logo bookings. Quarterly strategy review that checks the signposts in section 7 and the two conditional conditions (B10, B11). Excluded: new-logo bookings quotas for the horizontal team, which drove the old behaviour of signing any account, and discretionary discounting below the bundle.
Bottom-up coherence check. Compensation on net retention and module adoption sustains the bundle capability; the certification curriculum sustains implementation; the release gate sustains validation-ready engineering; the account P&L exposes whether implementation and connector reuse deliver the 79 percent gross margin and 17-month payback; those capabilities deliver the validated, integrated differentiation; that differentiation works only in companies large enough to need validation and integration and small enough that the enterprise platforms ignore them, which is the where-to-play; 560 such companies reach the aspiration. The chain holds in both directions.
Financial trajectory (fiscal 2028, from the value-driver-modeling case).
| Segment | Logos | ARR (M) | ARR per logo (k) | Gross margin | S&M (% ARR) | Contribution |
|---|---|---|---|---|---|---|
| Life-sciences vertical (bundle) | 560 | 94.1 | 168.0 | 79% | 28% | 51% (48.0M) |
| Horizontal mid-market, renewals and expansion only | 570 | 38.0 | 66.7 | 76% | 12% | 64% (24.3M) |
| SMB no-touch plan, 6,000 dollar minimum | 1,150 | 9.0 | 7.8 | 72% | 6% | 66% (5.9M) |
| Total | 2,280 | 141.1 | 77.7% | 22.3% | 55.5% (78.2M) |
Research and development at 24 percent (33.9 million) and general and administrative cost at 11 percent (15.5 million) give run-rate EBITDA of 28.9 million (20.5 percent) and a Rule of 40 score of 43 on 22.7 percent growth in the final year. Against the status-quo trajectory of 127.7 million of ARR and 7.0 million of EBITDA, option B adds 13.4 million of ARR and 21.9 million of EBITDA by fiscal 2028, and the ARR it adds is of a different quality: NRR above 115 percent rather than 103 percent. Path: fiscal 2026 ARR of 100.0 million (vertical 41.0, horizontal 46.0, SMB 13.0) and EBITDA of 8.0 million as SMB and horizontal run off faster than the vertical ramps; fiscal 2027 115.0 million (63.0, 41.5, 10.5) and 15.0 million; fiscal 2028 141.1 million and 28.9 million. Vertical logos run 300, 380, 470, 560 with ARR per logo of 96, 108, 134 and 168 thousand as accounts move onto the bundle. One-off costs of 9.5 million over three years: validation-ready release engineering 4.2 million, MES and LIMS connectors 1.8 million, restructuring of SMB support and the horizontal sales team including severance 2.6 million, certification and recruiting 0.9 million. Cash improves rather than worsens: three-year bundle terms billed annually in advance raise the share of vertical ARR billed in advance from 60 to 90 percent, adding about 14.1 million of deferred revenue at the fiscal 2028 run-rate (94.1 x 0.30 x 0.5, taking the average balance as half the annual billing).
7. Stage 8: activity-system map, fit tests, trade-offs, signposts
Activity-system map.
| Theme | Activities that deliver it | Reinforced by | Trade-off it requires |
|---|---|---|---|
| Validation-ready by default | Validation package (qualification documents, traceability matrix) produced by Sablecrest's own regulatory team for every release; fixed quarterly release cadence; audit-trail and electronic-signature controls on every record | Fixed cadence makes packages plannable; audit-finding benchmarks show which controls auditors test; bundle terms fund the regulatory team | No customer-specific customisation, because custom code breaks the package; no weekly release cadence for the horizontal base |
| GxP-certified implementation | 16-week fixed-scope implementation; one consultant per 12 accounts in steady state; certification curriculum; consultant runs the customer's first internal audit on the new system | Pre-validated configurations shorten implementation; certified connectors remove integration work; bundle price funds the consultant | No human-supported SMB tier: the people and the cost move from support to implementation |
| Certified integration layer | Connectors to the three ERPs, MES and LIMS, certified with each vendor and re-certified every release; referral programme with the ERP vendors' partner teams | Validation package covers the connector, so the customer validates once; connector data feeds the benchmarks | No accounts on uncertified ERPs, even when asked |
| Regulatory intelligence and benchmarks | Regulatory-change feed mapped to workflows; anonymised audit-finding benchmarks across customers; annual quality-leader council | Consultants and quarterly reviews collect audit outcomes; a larger vertical base improves the benchmarks | No non-regulated horizontal new logos: their data dilutes the benchmarks and their needs pull the roadmap |
| Bundle-based commercial model | Three-year six-module agreements; validation subscription at 12 to 15 percent above module list; compensation on net retention and module adoption; quarterly reviews on audit outcomes | Audit outcomes give the review its content; account P&L gives the compensation basis | No module pricing below the bundle; no new-logo bookings quotas for the horizontal team; no price-matching the low-cost tools |
Fit tests.
- First order (consistency): every activity points at validated, integrated, bundled service to mid-market life-sciences manufacturers. Two inconsistent activities were found: the horizontal team's new-logo bookings quota, which rewards signing any account, and the AI-copilot roadmap items scoped for horizontal use cases; both are removed.
- Second order (reinforcement): the fixed release cadence makes the validation package plannable, and the package makes each release sellable to a validated customer without a re-validation project; certified connectors shorten implementation, and shorter implementation lets one consultant carry 12 accounts, which is what makes the 28 percent sales and marketing ratio hold; consultants collect audit outcomes, and audit outcomes give the quarterly review its content and justify the validation subscription.
- Third order (optimisation): the position lets Sablecrest stop human-supported onboarding for 2,100 SMB accounts, stop custom development (14 open projects), stop maintaining nine uncertified integrations, and stop building horizontal features, which took about 30 percent of the roadmap.
- Interlock: the hardest activity for a rival to copy is the maintained validation package across every release combined with re-certified connectors; it is hard because it requires a regulatory team, a fixed cadence, re-certification with three ERP vendors every quarter, and a customer base whose audits feed the benchmarks. An enterprise platform would need a separate mid-market services model; a low-cost tool would need regulatory staff it does not have.
Trade-offs (what Sablecrest will not do, and what it costs). Stop new-logo selling to non-regulated horizontal accounts (3.7 million of new ARR a year foregone; the 51.2 million book is expected to run off to 38.0 million by fiscal 2028). Move 2,100 SMB accounts to the no-touch plan (about 7.0 million of ARR expected to leave over three years). Decline deals with companies over 2,000 employees, including two current customers acquired by larger groups that have asked for group-wide rollouts worth about 1.2 million of ARR combined, because winning them would pull consultants and the roadmap toward customers the enterprise platforms will fight for. Stop custom development (2.1 million a year of services revenue). Decline to price-match the low-cost tools on the horizontal base (an estimated 60 accounts will leave over it).
Signposts (conditions to keep under review, attached to the quarterly strategy review).
| Condition | Signal that it is failing | Measure and threshold | Response if triggered |
|---|---|---|---|
| B4 module consolidation | New vertical accounts plateau below four modules | Average modules and ARR per account by cohort at month six; below 3.6 modules or 165 thousand for two consecutive cohorts | Revisit bundle scope and validation subscription price before adding accounts |
| B8 consultant supply | Vacancies exceed 60 days | Open consultant roles over 60 days above 6 | Slow account additions; expand the certification intake |
| B10 run-off | SMB or horizontal ARR falls faster than modelled | Fiscal 2027 run-rate below 9.5 million (SMB) or 39.0 million (horizontal) | Accept, do not reinstate human support or horizontal selling; adjust cost base |
| B11 competitor response | An enterprise platform announces a validated mid-market edition below 100 thousand, or lost-bid reports show one implementing in under 20 weeks | Public announcements; lost-bid reports | Re-run war-gaming; consider pre-emptive three-year renewals with the top 50 vertical accounts |
| Growth trap (Porter) | Proposals to add verticals, tiers, geographies or enterprise deals | Any board paper proposing where-to-play expansion | Require a full WWHTBT before approval |
8. Stage 9: the choice record and hand-offs
Choice record, Sablecrest Software, approved by the board on the CEO's recommendation.
Framed choice: horizontal platform or vertical system of record, and on which basis.
Options considered and outcome:
| Option | Positioning base | Outcome | Deciding condition and evidence |
|---|---|---|---|
| A. Horizontal platform done well | None (operational effectiveness) | Rejected | A1 failed: SMB gross retention fell a further 2 points to an annualised 75 percent; 13 of 30 churned accounts went to spreadsheets, shared drives or a free tier |
| B. Quality system of record for mid-market life-sciences manufacturers | Needs-based | Chosen | B4 passed in pilot (7 of 10 expanded, 4.3 modules, 192 thousand ARR); B7 passed (79 percent gross margin, 17-month payback); B8, B10, B11 conditional with signposts |
| C. Best-of-breed supplier-quality point solution | Variety-based | Rejected | C1 failed: 20 percent of buyers on rival platforms would buy a standalone module, against a 35 percent pass mark; a single validated system of record cited as a requirement by 27 of 40 |
| D. Product-led long tail | Access-based | Rejected | D1 failed: segment sized at 112.8 million of addressable ARR against a 250 million pass mark |
Untested assumptions carried forward: B2 (segment spend growth at 12 percent or more; based on analyst forecasts and the ERP vendors' pipeline data, reviewed annually); B12 (the low-cost tools do not add validation packages or certified connectors; reviewed at each quarterly signpost review).
Decisions implied for the first 90 days: approve the 4.2 million validation-ready release programme and make the validation package a release gate from the next quarterly release; announce the no-touch SMB plan with a 120-day transition; move the 16 suitable implementation consultants into GxP certification and open the 32 external roles; sign three-year bundle renewals with the 40 existing full-bundle accounts on the new validation subscription; convert the 7 pilot accounts to three-year agreements; replace the horizontal team's new-logo quota with net-retention targets from the next quarter; hand the record to execution-planning for the programme and to kpi-architecture for the scorecard; hand the competitor conditions to war-gaming for the full scenario set; give the record to executive-communication for the board storyline.
Okafor's signed note on the record: "If module consolidation in new vertical cohorts sits below 3.6 for two cohorts, I will bring this back to the board rather than add accounts."