Licence fees are roughly a third of what a platform costs. The rest arrives on a schedule — front-loaded integration, a mid-period productivity trough, and a third-year repricing that happens precisely when you can no longer leave.
Martech business cases fail in a specific and repeatable way. They are not usually wrong about the licence. They are wrong about everything around it, and — more importantly — about when it lands.
The headline figures are now well enough established to plan against. Fully loaded, a martech platform costs roughly 2.5 times its licence fee, which means licence fees represent about a third of actual spending. CMOs underestimate true martech costs by somewhere between 40 and 60 per cent. One mid-market B2B example carried $850,000 in annual licence fees against a total annual stack cost of $2.1 million — a gap that was entirely predictable before anything was signed.
Those numbers get quoted often. The more useful observation is that the missing money does not arrive evenly. It arrives in three distinct waves, each landing on a different team, each visible to a different budget holder, and none of them appearing in the annualised figure that gets approved.
The cost curve, by year
Visual 1 — Where the unbudgeted cost actually lands
Phase | What arrives | Rough scale | How it appears in the business |
|---|---|---|---|
Year one — integration | Implementation and integration labour, plus ongoing maintenance capacity | Two to three times annual licence cost at implementation; maintenance equivalent to at least one full-time marketing operations resource, permanently | Unexpected change orders and a headcount request — rarely recognised as platform cost |
Years one to two — the ramp | Six to eighteen months of underutilisation while the organisation learns the platform | Paying full licence for partial capability; delayed campaigns, diverted team capacity, deferred revenue | A performance gap that gets attributed to the team, the market or the strategy |
Year three — the renewal | Recovery of the original discount, once switching costs are embedded | Uncapped, by construction, unless a cap was negotiated at signature | A renewal conversation in which the buyer's position is materially weaker than at purchase |
How to read it: Each phase is individually well known to whoever absorbs it. None of them is normally consolidated into a single number at the point of approval, which is why the same surprise recurs across organisations and across categories.
The third wave is the interesting one
Year-one integration overrun is familiar enough to be priced by anyone who has run a migration. The adoption ramp is understood, if optimistically. The third-year repricing is different, because it is usually read as vendor opportunism when it is closer to arithmetic.
Consolidation and displacement discounts are priced against the buyer's freedom to walk away. At signature that freedom is real, which is what the discount reflects. Eighteen months later the integrations are built, the workflows are rewritten, the team is trained, the historic data lives in the platform — and the freedom the discount was priced against no longer exists. The renewal is not a new negotiation. It is the first negotiation, conducted at the true price.
Leverage transfers to the vendor at the moment implementation completes — which is also the moment the programme is declared a success and the people who negotiated the contract move on to something else.
The practical consequence is that the only time to address year three is at signature. Price protection, renewal caps, defined exit assistance and data-extraction terms cost nothing to ask for while a deal is being won and are close to unobtainable once the platform is embedded. Organisations that discover the increase without a cap in place, as the reporting on this puts it, discover it too late.
Consolidation does not reliably remove anything
There is a second assumption doing quiet damage in these business cases: that replacing five tools with one removes five costs.
It frequently does not. Some 82.7 per cent of organisations continue to employ alternative products after consolidating — the displaced tool retained by a team with a use case the suite does not cover, or a contract with eighteen months left, or an integration nobody wants to unpick. Meanwhile stack utilisation sits at around 49 per cent, meaning the average organisation is paying roughly 2.5 times licence for a platform it uses about half of.
Put those together and the consolidation case that promised subtraction often delivers addition. The suite arrives, the point solutions persist, and the saving that justified the programme is recovered from a line that was never real.
This is not an argument against consolidating. It is an argument that the business case must include a decommissioning plan with named owners and dates, and that a consolidation projected to save money while retaining the displaced tools is not a saving — it is a purchase.
What a defensible martech business case contains
A three-year total cost line, not an annual licence line. Licence is roughly a third of it; build the other two-thirds explicitly rather than as contingency.
Implementation costed at two to three times annual licence, and ongoing maintenance costed as at least one permanent marketing operations resource.
An explicit ramp assumption. Six to eighteen months of reduced output, stated as a number, with the campaigns that will not run during it named.
Renewal price protection negotiated at signature. A cap, a defined uplift ceiling, and contractual exit assistance including data extraction in a usable format.
A decommissioning schedule with owners and dates for every tool the platform replaces, and a mechanism for what happens if a team refuses to give one up.
A utilisation target with a review date. Against a 49 per cent baseline, buying more capability than the team can absorb is the default outcome, not the risk case.
What this changes
None of these figures is a secret, and none of them requires new research to apply. What they require is a business case built on a three-year sequence rather than an annual average — because the annualised number is wrong in year one, wrong in year two, and wrong again in year three, in different directions each time, which is why it survives review.
For marketing operations leaders heading into planning season, the single highest-return change is the least technical one: move the renewal conversation to signature. Everything else on the list is a forecasting improvement. That one is a transfer of leverage, and it is only available once.
Sources and method. A MarketingHubMedia original. Cost figures — total cost of approximately 2.5 times licence fees when hidden costs are fully loaded; licence fees representing roughly one-third of actual spending; CMOs underestimating true martech costs by 40 to 60 per cent; implementation and integration running at two to three times annual licence cost with ongoing maintenance equivalent to at least one full-time marketing operations resource; a six-to-eighteen-month adoption ramp; third-year renewal recovery of consolidation discounts once switching costs are embedded; 82.7 per cent of organisations still employing alternative products after consolidation; 49 per cent stack utilisation; and the mid-market example of $850,000 in annual licence fees against $2.1 million in total annual stack cost — as reported by MarTech. The three-phase cost-curve framing, the leverage-transfer argument and the business-case checklist are MarketingHubMedia's own analysis. Figures are reported as published and not independently re-verified. Journalism, not procurement advice. Corrections will be made openly on this article.
