The Three Options on Every Whiteboard

At some point every demand-gen leader draws the same diagram. On the left, a target account list. On the right, booked meetings. In between: enrichment, intent scoring, asset generation, a landing page layer, and attribution.

There are three ways to fill the middle.

Build it: your engineers assemble the pipeline from APIs and internal glue. Stitch it: you buy five point tools and integrate them. Or adopt it: one platform that ships the middle as a single composed system.

Most teams price these options on licence cost. That is the least informative number in the comparison.

What Building Actually Costs

The build option looks attractive because the components seem commoditised. Enrichment APIs exist. Page builders exist. An engineer can wire a prototype in a fortnight.

The prototype is not the product. The product is the prototype plus retry logic, rate-limit handling, schema drift management, deliverability plumbing, monitoring, and a person on call when a data provider changes their response format on a Friday. The fortnight becomes a quarter. The quarter becomes a permanent engineering allocation, because the pipeline is never finished, only currently working.

The honest build cost is a headcount line, forever. For a company whose product IS demand infrastructure, that can be worth it. For a company whose product is something else, it is a tax on the roadmap.

What Stitching Actually Costs

Stitching is the default, which is why the average B2B stack sprawls into dozens of tools. Five contracts for the demand middle alone is normal.

The visible cost is five licences. The real costs sit underneath.

Integration debt: each tool speaks its own schema, so every pair of tools needs translation glue, and the glue breaks whenever any vendor ships a breaking change on their own release cycle. Five tools means the failure surface is not five things; it is the connections between them.

Data fragmentation: the enrichment record lives in one system, engagement lives in another, outcomes live in a third. Nobody holds the whole picture of an account, so attribution becomes archaeology.

Renewal drag: five separate negotiations, five separate price rises, five separate procurement cycles competing for the same budget line.

And the quiet one: pace. Every campaign idea that spans two tools needs an integration conversation first. The stack starts dictating what the team can imagine.

What Adopting Actually Costs

The composable platform option has real costs too, and pretending otherwise would be the kind of claim this article exists to avoid.

There is a migration, even a well-sequenced one. There is vendor concentration: one platform matters more to you than any single point tool did. And there is a fit question: a platform's opinionated pipeline has to actually match how your team works, which is something to verify in a demo against your own list, not in a feature matrix.

What you get back is structural. One schema from enrichment to attribution, so the account record is whole. One integration surface instead of the pairwise mesh. One renewal. And a pipeline where the stages were designed to feed each other: enrichment output IS the personalisation input, rather than being translated into it through glue.

The Decision Framework

Strip it to four questions.

First: is demand infrastructure your product or your plumbing? If it is the product, build. If it is plumbing, do not staff it like a product.

Second: count the handoffs. How many times does an account record cross a system boundary between list upload and booked meeting? Each crossing is glue you maintain and a place data drops. If the answer is more than two, stitching is already taxing you.

Third: who owns the whole record? If answering "what happened with this account" requires three logins, attribution will always be an estimate.

Fourth: what does your platform team NOT ship because they maintain the stack? This is the cost nobody invoices. It is usually the largest number in the comparison, and it never appears in it.

Where the Numbers Live

We wrote the long version of this argument, with the cost model itemised line by line and a sequencing plan for consolidating without a big-bang migration, as The Consolidation Imperative whitepaper. If this decision is on your desk this quarter, it is written to be forwarded to the person who signs.