You use half your martech because buying is easier than adopting. Every gap triggers a new licence, and the last one never got fully rolled out. The fix is not another tool. Audit the stack against real usage, cut the platforms nobody runs, and move the savings into the people who run the rest.
Open your billing page and count the logins nobody has touched this quarter. Then count the features switched off inside the tools you do use. There is a second stack hiding inside the one you pay for, and most of it is dark.
Why you only use half the stack
Every gap gets its own tool. A campaign need appears, someone finds a platform that solves it, and a licence lands. The tool works. What rarely happens is the full rollout: the training, the documentation, the habit of using it end to end. Then the next gap appears, and the cycle repeats before the last tool was ever adopted.
So capability piles up faster than anyone can absorb it. The average team runs 20 to 29 tools, drawn from a market that grew from 150 products in 2011 to nearly 10,000 by 2022 (2X). Nobody set out to build a sprawling stack. It accreted, one reasonable purchase at a time.
The result is a stack you own but do not operate. The features are there. The seats are paid. The knowledge to run them is not, because attention moved on before adoption caught up.
The evidence: half the capability sits idle
Marketers use 49% of their martech capability (Gartner, 2025). Half of what you bought is dark. Not broken, not missing, just never switched on.
That gap now shows in the budget. Martech’s share of marketing budget fell to 22% in 2025 from 30% in 2023, driven by doubt about return (Gartner via CMSWire). CMOs are not anti-technology. They are watching the last few licences fail to prove themselves, and pulling back.
I have run the other version of this. Building Cisco’s inbound demand-generation engine to a 628 million dollar pipeline was not a buying exercise. It was making a stack we already had actually run, end to end, by the people paid to run it. The pipeline came from adoption, not procurement.
Is this you?
Five quick checks. Answer each yes or no.
- Can you name a tool you pay for that nobody has logged into this quarter?
- Do two of your platforms do broadly the same job?
- When you hit a new marketing need, is your first instinct to buy rather than to check what you already own?
- Did your last platform get a real rollout, with training and documentation, or just a login?
- Could you list, from memory, which features of your main tool your team never uses?
Three or more uncomfortable answers means you are paying for capability you do not run.
What underutilisation actually costs
The cost is invisible because it is already paid. Underutilisation can cost a $250 million company up to $4 million a year in capability that sits idle (Gartner, 2025; and 2X). No invoice says “unused features”. The money left the building the day you signed.
It shows up three ways. As duplicate spend, where two tools cover the same job and you pay twice. As firefighting, where a task that a switched-off feature would handle gets done by hand every week. And as capped AI, because the models you want to run depend on data and workflows sitting inside tools nobody operates fully.
None of it appears as waste. All of it is money you already spent and never got back.
Three ways to get value from what you own
The move is not another purchase. It is the opposite.
Audit the stack against actual use. List every tool, its cost, and its real login and feature usage. Not what it could do. What your team does with it. The redundancies and the dark features become obvious fast, and the audit itself usually pays for the effort.
Cut the redundant platforms. Where two tools do one job, keep the one your team actually runs and drop the other. This is where the buy-a-tool reflex gets unwound, and where the savings that fund the next move come from.
Move the savings into enablement. Training, documentation, and a culture of using what you own. The historic 10/90 rule holds: spend 10% on tools and 90% on the people who run them. Adoption, not acquisition, is what turns a paid licence into revenue. Most of this also traces back to a question the sprawl hides: who actually owns your martech stack.
That is the shape of it. The full sequence, with owners and what to measure, is the playbook. Get the full martech utilisation playbook.
Buy last, not first
Sprawl is not a technology problem. It is a reflex: reach for a purchase order when the answer is already sitting, unused, in the stack you have.
The teams pulling ahead are not the ones with the most tools. They are the ones who run what they own, cut what they do not, and spend on the people who operate the stack. That is a cheaper path than the next licence, and a faster one. The value is already paid for. It is waiting to be switched on.
Sprawl feeds the layer below it, where every disconnected tool becomes its own store of record. If your CDP has quietly become one more silo, that is CDP disillusionment, and it is the next obstacle down.
Go deeper on customer data maximization
Three ways forward. Pick the one that fits where you are.
- Get the playbook. Practical notes on turning the customer data you already own into revenue, straight to your inbox. Join the newsletter at the foot of this page.
- Take the assessment. Score your customer data maximization in four minutes and see your top revenue blockers. Start the assessment →
- Book a meeting. Bring your data problem. Leave with a prioritised fix, not a platform pitch. Book a call →
Post 16 of 25 in the Customer Data Maximization series. Previous: Channel proliferation and self-service. Next: CDP disillusionment and the composable answer.
Frequently asked questions
How much of their martech do marketers actually use?
Marketers use 49% of their martech capability, so roughly half the features you pay for sit idle (Gartner, 2025). The gap is rarely a tooling problem. It is an adoption problem: the last platform was never fully rolled out before the next one arrived.
How much does martech underutilisation cost?
Underutilisation can cost a $250 million company up to $4 million a year in unused capability (Gartner, 2025; and 2X). That value is already paid for. It sits in features nobody was trained to run, which is why closing the gap is a quick win rather than a new investment.
How many tools does a typical marketing team run?
The average team runs 20 to 29 tools, drawn from a market that grew from 150 products in 2011 to nearly 10,000 by 2022 (2X). Every gap gets its own login, and few of them overlap cleanly, so capability duplicates and adoption spreads thin.
Why is martech's share of marketing budget falling?
Martech's share of marketing budget fell to 22% in 2025 from 30% in 2023, driven by doubt about return (Gartner via CMSWire). CMOs are questioning whether the next licence pays back, because the last few did not get used enough to prove it.
What is the fastest way to get more from your martech?
Audit the stack against actual use. Cut the platforms nobody runs, then move the savings into enablement: training, documentation, and a habit of using what you own. The historic 10/90 rule holds: spend 10% on tools and 90% on the people who run them.