Revenue: the eleven levers that decide who keeps the saving

The order of operations

Three of these levers are about pricing (R1, R8, R11), two are about keeping what the firm has (R4, R3), one is the condition-three deliverable (R2), and five are Compound-stage growth that only exists once the first six are done (R5, R6, R7, R9, R10). The order is not optional. A firm that reaches for down-market entry before its pricing is converted has built a cheap channel for giving work away.

Evidence in this group is better for the pricing levers, because whole sectors have repriced and said so, and worst for the Compound levers, because almost nobody has reached them. The cards say which is which.

R1. Pricing model conversion

What it is. P9 converted what renewed in the first cycle. R1 carries the rest of the book across the following year and adds the tiering that a fixed price makes possible: a base tier for the platform-delivered service, a standard tier with human review and advisory time, a premium tier (R11). The end state is a price list, which a services firm has never had, and a recurring-revenue share that a buyer can audit.

Line it moves. Revenue, and the scorecard's first line, recurring share. Also the flow-through of every D-group saving.

Typical range and evidence. Sector-level: in small-firm US accounting, fixed fees are the dominant model at 54 percent of firms and hourly billing is under 4 percent, per a billing-platform vendor's 2025 benchmark.¹ At the top of the market, the US head of one of the Big Four told the Financial Times in March 2026 that the firm will sell AI tax and consulting tools by subscription without an hourly-billing consultant.² Among consultants, a 2026 fee survey's preliminary results put the share billing primarily by the hour at 30 percent, down from 36 percent two years earlier.³ What no public source measures is the price a given firm achieves against its hourly-equivalent, which is the number that matters. Grade: Medium (sector); the firm's own conversion price is measured from P2.

How to run it. The price per service line comes from P2's "what would they pay for the outcome" exercise, cross-checked against realization: a firm realizing 85 percent of recorded hours has been giving 15 percent away and the fixed price should recover some of it, not all. Tiering follows the unit table: platform-delivered units in the base, review and advisory hours in the standard, named-senior access in the premium. Grandfathered discounts end at renewal. Clients who insist on hourly terms keep them at a premium to the fixed equivalent.

Conditions. P9 done, so the first cycle has set the pattern. D3 live, so the base tier has a quality story.

Kill criteria. Price per unit against hourly-equivalent below 90 percent at the second renewal cycle (the conversion is a discount in disguise). Hourly share of revenue above 25 percent at the end of year two.

Metrics. Year one: recurring share; price per unit versus hourly-equivalent; share of clients on a tier. Year three: hourly share immaterial; tier mix.

Horizon and stage. Convert; day 100 to year two.

Across the five firms. A: retainers become platform subscriptions with an advisory tier. B: already fixed; R1 is the tiering. C: per-matter fixed fees first, then a subscription for the recurring matter types. D: permits and renewals as subscriptions; one-off projects stay fixed-fee. E: already outcome-priced; R1 does not apply.

R2. The platform-delivered service line

What it is. Condition three. One service the firm sells, delivered end to end on the platform, client-facing (a portal, continuous monitoring, a turnaround the client can see), billed as recurring, live with paying clients inside twelve months of close. Not back-office automation that the client never notices. The client experiences a different product, and that is what the twelve-month test is for.

Line it moves. Revenue and mix. More than that, it is the first evidence a buyer at exit will accept that the firm has become something other than a services firm.

Typical range and evidence. The pattern exists and the outcomes do not. A legal platform acquired a 300-person legal services provider for €42 million in July 2025 and launched a fixed-fee, platform-plus-people offering with a named Fortune 500 first client within weeks; its results are its own press material.⁴ A compliance platform acquired a broker-dealer compliance consultancy in 2025 to offer "an end-to-end solution"; no results published.⁵ Grade: Weak (no published outcomes); the firm's own line is measured from launch.

How to run it. P6 picked the service line: the one whose units are mostly "ready" and whose clients are on converted pricing. The critical path is P6's build list, D4's frameworks for that line, D3's sampling plan, and the client-facing surface (which a platform company has and a services firm never did). Launch with the firm's most committed clients from P3's map, under a named-senior premium tier (R11) so the client gets the platform and the person. The metric is clients live on the line and recurring revenue from it, not features shipped.

Conditions. P6's build inside nine months. P9 pricing on the launch clients. D3 at 100 percent sampling at launch.

Kill criteria. No paying client on the line at month twelve (condition three failed; the register's largest kill). Client churn on the line above the firm's average in its first year.

Metrics. Year one: clients live; recurring revenue from the line; its error rate. Year three: share of total revenue delivered through platform lines.

Horizon and stage. Convert; year one.

R3. Scope expansion within the client

What it is. A services firm's expansion is limited by its hours: a client who wants three more jurisdictions covered needs three more analysts' worth of time. A platform-delivered firm expands at marginal cost: more regulations, more entities, more frameworks, more products under the same client relationship, each one a configuration rather than a hire. This is the lever that turns the book into net revenue retention above 100 percent.

Line it moves. Revenue (net retention).

Typical range and evidence. Not measured in services conversions. The nearest analogue is the software edition's expansion lever, where a sponsor's portfolio data showed revenue per customer-success manager rising as expansion plays were run on signals rather than relationships.⁶ The mechanism transfers; the number does not. Grade: Weak (argued from analogue); the firm's NRR is its own.

How to run it. Map each client's current scope against the platform's full scope. The gap is the expansion list. Price the increments as tiers or add-ons (R1). Have the relationship holder (P3) sell them, because the client bought the person before it bought the platform.

Kill criteria. NRR below 100 percent at year two (the book is shrinking faster than it expands; see R4).

Metrics. Year one: expansion revenue from existing clients; NRR. Year three: NRR; scope per client against platform scope.

Horizon and stage. Convert; years one to two.

R4. Book retention through conversion

What it is. Condition five, run as a lever. Client-level revenue tracked against the close baseline, every month, with a named owner per client above one percent of revenue, and intervention targeted where it works. The software edition carried a rule from the retention literature that transfers unchanged: the clients most at risk are not necessarily the ones most worth intervening with, because some will leave whatever you do and some would have stayed anyway; target the ones whose decision the intervention changes.⁷

Line it moves. Revenue (gross retention), and the whole thesis: a conversion that hits every margin target and loses a fifth of the book has failed.

Typical range and evidence. No independent evidence on client attrition after an ownership change in professional services; the first study is in progress and has not reported.⁸ The targeting rule is Strong and carried. Grade: Strong (targeting rule) / Weak (sector attrition data); runs on the firm's own baseline.

How to run it. P3's map gives the owners. P8 gives the baseline. The monthly report is client-level revenue against baseline, flagged where it falls, with the owner's note. Interventions are a senior visit, a premium-tier offer (R11), or a scope conversation (R3), chosen by the owner; they are not a discount, because a discount at conversion is P9 undone.

Kill criteria. Client-level revenue below 95 percent of the close baseline in aggregate at any quarter in year one, or below 90 percent at any point thereafter. Any top-ten client lost.

Metrics. Monthly: book retention against baseline (scorecard line five). Year one and three: the same, plus the share of revenue held by people hired after close (P3's year-three metric).

Horizon and stage. Convert; day 100 to year three.

R5. Throughput as revenue

What it is. Faster delivery sold. A platform-delivered unit that took three weeks and takes three days can be sold as a premium turnaround, or the freed capacity can be sold as more engagements without more hires. Either way the cycle-time gain from D1 and D6 becomes revenue rather than slack.

Line it moves. Revenue.

Typical range and evidence. Not measured. In one archetype the client's own economics make the case: in revenue cycle, more than half of denied private-payer claims are reported to be overturned on appeal, so capacity to appeal more claims faster is revenue for the client and for a firm paid on collections.⁹ Grade: Weak (argued); Medium in archetype E.

How to run it. Measure cycle time per unit type before and after (D6). Offer the faster turnaround as a tier where clients value speed (filings against deadlines, claims against timely-filing windows). Where they do not, book the capacity as new engagements at the current team size.

Kill criteria. Cycle time not improved by half for the converted unit types at year one (D1 is not working as claimed).

Metrics. Year one: cycle time per unit type; revenue from turnaround tiers; engagements per delivery FTE. Year three: engagements per delivery FTE.

Horizon and stage. Convert; years one to two.

R6. Down-market entry

What it is. The clients a services firm could never serve profitably: the small manufacturer with one product line and one regulation, the ten-person company that needs a return and nothing else, the clinic with a hundred claims a month. At manual cost they were turned away or served badly. At platform cost they are a segment, and the firm can enter it with a base-tier product and a lighter sales motion (R7).

Line it moves. Revenue and new logos; also the mix, because down-market revenue is smaller per client and more numerous.

Typical range and evidence. The argument is the central one of the services-as-software thesis: software reaches customers who could never afford the human version. It has been made by venture investors as a market-sizing claim, not measured as an outcome at any converted services firm.¹⁰ Grade: Weak (argued).

How to run it. Only after R1 and R2: the base tier must exist and be platform-delivered, or the down-market client is served manually at a loss. Define the segment by the units it needs (usually the "ready" ones only). Price from cost per unit (D7) with a software-like margin. Sell through R7's motion, not through partners.

Kill criteria. Cost to serve a down-market client above its annual revenue at year two. Down-market churn above twice the core book's.

Metrics. Year two: down-market clients; revenue; cost to serve; CAC payback (scorecard line seven). Year three: down-market share of new logos.

Horizon and stage. Compound; years two to three.

R7. New-client acquisition at software economics

What it is. A services firm acquires clients through partners' networks, referrals and conference rooms. The cost is invisible because it is partner time, and the ceiling is the number of partners. A platform-delivered firm can acquire through marketing, inside sales and a product the prospect can see, with a measurable acquisition cost and a payback period. This lever builds that motion, and its evidence carries a headwind from the software edition: search-driven demand generation has weakened as answer engines absorb clicks, so the motion has to be built for the market as it is now.¹¹

Line it moves. Revenue; sales and marketing cost (which, for the first time in the firm's life, appears as a line).

Typical range and evidence. The headwind is Strong and carried; the fix is Weak and carried. For a services firm the additional fact is that it starts from near zero in marketing, so the comparison is not against a prior CAC but against partner time that was never costed. Grade: Strong (headwind) / Weak (fix).

How to run it. Cost partner selling time at close, so there is a baseline. Build the inside motion around R2's line, because it is the thing a prospect can be shown. Proposal and RFP automation (O1) carries the volume. Measure CAC payback from the first quarter and compare it to the costed partner motion.

Kill criteria. CAC payback above 24 months at year two for the platform-delivered lines.

Metrics. Year one: costed partner CAC baseline; inside-motion CAC and payback. Year three: share of new logos from the inside motion; blended CAC payback (scorecard line seven).

Horizon and stage. Compound; years one to three.

R8. Episodic to continuous

What it is. A services firm's work is episodic: an annual assessment, a filing at year-end, a project when the regulation changes. The client pays for the episode. A platform that holds the client's data can monitor continuously, flag changes as they happen, and deliver the episode's output as a by-product of an always-on service. The pricing follows: a subscription for monitoring, with the episodes included. This is the lever that answers the critics' third reason, that short contracts leave no lock-in: a monitoring subscription is renewed because switching means losing the history.¹²

Line it moves. Revenue; recurring share; retention (a subscription with history is stickier than a project).

Typical range and evidence. The move is visible at the top of the market in the Big Four subscription statement,² and in compliance it is the stated direction of platform-plus-assessor combinations (continuous compliance rather than periodic engagement).⁵ Outcomes unpublished. Grade: Medium (direction) / Weak (outcomes).

How to run it. Identify the episodic engagements whose inputs the platform already holds after R2. Offer the monitoring subscription at renewal, with the annual episode included, priced above the episode alone and below the episode plus a second engagement. The client buys fewer surprises; the firm buys a recurring line.

Kill criteria. Conversion of episodic clients to subscription below 30 percent after two renewal cycles.

Metrics. Year one: episodic revenue converted to subscription. Year three: recurring share from monitoring subscriptions; renewal rate on them versus episodic clients.

Horizon and stage. Convert; years one to two.

R9. Cross-sell across the roll-up

What it is. The second acquired firm's clients offered the first firm's services, and vice versa, on one platform (B3). In a conventional services roll-up this is the synergy that rarely appears, because each firm's partners sell their own work. On a shared platform with a shared client record, the cross-sell is a configuration and a conversation rather than a referral between partners who do not know each other.

Line it moves. Revenue (net retention across the group).

Typical range and evidence. Not measured in this category. The conventional roll-up evidence is cautionary: acquisition-led growth is visible to a buyer, and the one public example of an acquisition-led platform reporting organic growth showed 3 percent against 126 percent total.¹³ Grade: Weak (argued); the cautionary comp is Strong.

How to run it. Only after B3 has put the acquired firms on one platform and one client record. Map each client's scope against the group's scope (R3 at group level). Compensate the relationship holder for group revenue, not firm revenue (O7).

Kill criteria. Cross-sell revenue below 5 percent of the acquired firm's book two years after its conversion (the platform is shared but the selling is not).

Metrics. Year two onward: cross-sell revenue per acquired firm; group NRR versus firm NRR.

Horizon and stage. Compound; years two to three.

R10. Data and benchmark products

What it is. The aggregate of the work: what regulations clients are exposed to, what suppliers fail which screens, what payers deny for what reasons, what filings take how long. Sold as benchmarks, indices or regulatory content, inside the rights P4 established and the governance X3 built. The software edition graded this lever Weak and it stays Weak here: it is argued from the pricing power of companies that own regulatory content, not measured at any converted services firm.

Line it moves. Revenue.

Typical range and evidence. Argued. Grade: Weak.

How to run it. Not before year two, and not before P4's data-rights share is a majority of the book. Start with a benchmark given to clients as part of the premium tier (R11), which tests demand without a sale. Sell only once the governance (X3) can show a client exactly what of theirs is and is not in the product.

Kill criteria. P4's data-derivative rights below 50 percent of revenue (the product cannot be built honestly).

Metrics. Year three: data-product revenue; share of clients consenting.

Horizon and stage. Compound; years two to three.

R11. Premium human tier

What it is. The senior people's new role, priced. When the platform delivers the base tier, the partner's judgment, the named relationship and the accountability a client wants when something goes wrong become a product in their own right, sold above the platform tier rather than bundled into hours. This is the lever that answers the critics' fourth reason, that inefficiency is the product because clients pay for accountability and someone to call: the playbook agrees, and prices it.¹²

Line it moves. Revenue; and partner retention, because it gives the people from P3's map a role with a price on it rather than a role that has been automated.

Typical range and evidence. Not measured. The analogue is in every tiered software price list, where a named account team and an SLA sit in the top tier. In services the evidence is structural: fixed-fee engagements in accounting often still carry a billable-hour metric inside them, which is the premium tier unformed.¹ Grade: Weak (argued).

How to run it. Define the premium tier as access (a named senior, response times, a quarterly review) rather than as hours. Price it as a share of the client's spend, not as a rate card. Put the relationship holders from P3 in it by name, which is also how O7 pays them.

Kill criteria. Premium tier attach below 20 percent of the converted book at year two (clients do not value the named senior as much as the firm believed, which is a P3 finding as much as an R11 one).

Metrics. Year one: premium attach; revenue per premium client. Year three: premium share of revenue; partner retention among premium-tier owners.

Horizon and stage. Convert; year one.

What the group adds up to

The scorecard carries four of this group's outcomes: recurring share (R1, R8), book retention (R4), net retention (R3, R9) and CAC payback (R6, R7). A sponsor reading them together is reading the answer to the Fortune objection's four reasons: pricing converted (reason two), lock-in built (reason three), accountability priced (reason four), and, through R2, the firm delivering through its own platform rather than renting labor on the client's terms (reason one). If those four lines move, the re-rating case is made on evidence. If they do not, the delivery margin is real and the firm is a better services business, and X6 says how to sell that honestly.