Enterprise SaaS companies keep adding to the martech stack, yet the return on that spend keeps falling. To understand why, we measured the observable marketing technology footprint of 190 enterprise SaaS companies and scored each martech stack across seven dimensions, then set the results against the published research on how that technology is bought and used. The pattern that comes back is consistent and uncomfortable: the enterprise martech stack is built to measure, not to convert. This report lays out the evidence, the size of the gap, and the specific sequence of work that separates the small group of high performers from everyone else.

The stack has changed

The enterprise martech stack did not grow deliberately. It accreted. Every new channel, every new tactic, every new vendor category added a layer, and almost nothing was ever removed. The Chiefmartec landscape has grown from roughly 150 products in 2011 to 15,384 in 2025, a hundredfold expansion in fifteen years, and it is still climbing about 9 percent a year even as the market supposedly consolidates. For the enterprise buyer, the practical result is a marketplace with far more overlap than clarity, with dozens of vendors in every category from marketing automation to CRM, each promising the same lift.

The growth has not been evenly spread across the martech stack, and that unevenness is the first clue to the problem. The categories that expanded fastest are the cheapest to adopt and the easiest to justify: web analytics, tag management, content tooling, and a long tail of point solutions that each solve one narrow task. The categories that would actually change outcomes, a unified data layer, disciplined experimentation, revenue-grade attribution, grew far more slowly, because they are harder to buy, harder to deploy and slower to show a quick win. The enterprise martech stack that resulted is shaped like an inverted pyramid, broad and heavy at the measurement layer and thin and fragile at the layer that turns measurement into money.

Spending followed the sprawl. Marketing technology now accounts for roughly a fifth of the total marketing budget, a line item large enough to demand board-level justification, yet one that most organisations cannot tie back to revenue. That is the tension every CMO now carries into budget season. The stack is not under-funded; it is under-orchestrated. The question is no longer whether a company owns enough marketing technology. It is whether the technology it already owns is arranged to produce growth, and for the large majority of the companies we studied, it is not.

For the person holding the budget, the real problem is coherence rather than choice. Every new purchase adds a login, a data source and an integration to maintain, and subtracts from the time the team has to make any single tool work. A martech stack with forty tools that do not talk to each other is not more capable than one with twenty that do. Past a certain point, each addition raises the coordination cost faster than it raises output. Categories such as lead generation and sales automation illustrate the pattern well: buyers face a dozen credible options, adopt two or three, integrate none of them fully, and end up paying for capability they never switch on. That is the shape of the modern martech stack, and it is the reason the rest of this report exists.

The efficiency gap

Read the martech stack as a value chain, moving from collecting data, to unifying it, to acting on it, to governing it, and the enterprise SaaS pattern becomes unmistakable. The front of the chain is crowded. The back of it is empty. We measured how many of the 190 companies run at least one detectable tool at each stage, and the drop-off is steep and consistent.

Almost every company has analytics and a tag manager, and a modern content platform is close behind. These are the tools that answer the question, what happened. They are cheap to justify, quick to deploy, and by now a default. Two-thirds of enterprise SaaS marketing organisations have effectively solved measurement, and for many teams that is where the martech stack quietly stops growing in any meaningful way.

Then the chain thins out fast. A customer data platform, the system that turns scattered analytics into one activatable view of the customer, appears on barely a quarter of companies. Experimentation and conversion optimisation tooling sit at 31 percent. Conversational tooling that turns intent into a live sales conversation reaches only one company in five. The capabilities that move measurement toward revenue are precisely the ones most enterprises have not bought, or have bought and never fully deployed. Each missing stage does not just remove its own value; it strands the value of the stages before it, because analytics with nowhere to flow is a report, not a system.

This is the efficiency gap made concrete. When 36 percent of companies run a complete measurement foundation but no customer data platform and no experimentation layer, the missing spend is not on dashboards. It is on the machinery that would turn those dashboards into pipeline. The martech stack is optimised to report on growth it is not built to produce. Notice, too, that consent tooling now sits at 47 percent, higher than marketing automation at 42 percent. Enterprises have staffed the privacy layer faster than the growth layer, which tells you something about how these decisions get made: the tool that reduces obvious risk is easier to approve than the tool that creates non-obvious upside. Everything that follows in this report, the integration tax, the AI gap, the attribution failure, is downstream of this one shape.

The integration tax

Owning the pieces of a martech stack is not the same as connecting them, and the gap between the two is where most enterprise value leaks away. On the Efficiency Index, Integration is one of the weakest dimensions, with the average company scoring under half of the available points. The outside research shows exactly what that disconnection costs.

Integration is the stage where a martech stack either compounds or leaks. A customer data platform wired to analytics, automation and advertising turns every touchpoint into a usable signal that the next system can act on. The same tools sitting side by side, unconnected, turn every touchpoint into a separate silo that has to be reconciled by hand, if it is reconciled at all. Our data finds a genuine unification layer, a CDP or a mature tag-management and reverse-ETL spine, present on fewer than half of companies. The rest run their tools in parallel rather than in concert, and pay for the difference in analyst hours, duplicated records and decisions made on partial data.

Forrester finds that 41 percent of B2B marketers are held back by siloed data and a lack of integration across systems, and independent analyses put the ROI penalty for disconnected marketing data at 15 to 20 percent. Separate research on knowledge workers finds that staff lose hours every week chasing data across disconnected tools, a cost that never appears on the martech invoice but shows up in every delayed campaign and every report that has to be rebuilt from scratch. The integration tax is real, it is quantified, and it compounds the more tools you add without connecting them.

This is the mechanism behind the efficiency gap. A company can pass every individual tool audit and still fail as a system, because value in a martech stack lives in the connections, not the components. The enterprises that keep buying more tools without integrating them are not closing the gap; they are widening it, and paying twice along the way, once for the software and again for the growth it quietly fails to produce. The practical consequence is that integration, not acquisition, is the highest-return work available to most enterprise marketing teams right now, and it is work that a focused marketing consulting engagement can usually sequence in a single quarter.

AI: adoption vs reality

No category is louder in the 2027 sales deck than AI. In the observable martech stack, it is the quietest. AI Readiness is the single weakest dimension on the Efficiency Index, scoring just 27 percent of its maximum across the sample. The gap between how much AI is discussed and how much is actually deployed in enterprise marketing is the widest of any capability we measured.

There is an irony worth naming plainly. The sample is dominated by enterprise SaaS companies, many of them selling AI-powered products, yet the AI-enabled marketing capability we can observe in their own stacks, personalisation, conversational engagement, machine-driven experimentation, is thin. The distance between the AI in the pitch and the AI in production is not a single marketing team failing; it is an industry-wide one, and the buyers can feel it. Gartner reports that 45 percent of martech leaders say vendor-supplied AI agents failed to meet the business performance they were sold. Our own review of AI marketing tools found the same split between an impressive demo and a disappointing deployment.

The lesson is not that AI is overhyped in the abstract. It is that AI cannot compensate for a missing activation layer. An AI that has no unified customer data to reason over, no experimentation loop to learn from and no automation to act through is a demo, not a system, and it will inherit every weakness of the martech stack it is bolted onto. The enterprises getting real value from AI in marketing are the small minority that already built the activation and integration layers underneath it, so the AI has clean data to work with and a mechanism to act. For everyone else, buying AI on top of a stack that was built to measure simply adds a fast, confident way to produce the same disconnected outputs. The order of operations matters here more than almost anywhere else: data and activation first, intelligence on top, never the reverse.

Search behaviour is compounding the pressure. Buyers increasingly begin their research inside AI answer engines rather than a list of blue links, which means a brand that is invisible to those engines is invisible at the very start of the journey. That is a different discipline from classic optimisation, and it is why AI search and answer-engine optimisation has moved from a nice-to-have to part of the activation layer itself. The martech stack that cannot be found by an AI is missing demand it never even sees.

Attribution is broken

Everything upstream, the missing activation layer, the integration tax, the un-deployed AI, converges on a single failure that CMOs feel most acutely: they cannot prove what their marketing produced. This is the point where a technical shortfall becomes a boardroom problem, because a function that cannot measure its own output cannot defend its own budget.

The numbers are stark. Barely 36 percent of marketers say they can accurately measure marketing ROI, which means roughly two in three cannot. 82 percent struggle to connect data across the customer’s touchpoints, the exact capability a customer data platform and a connected martech stack exist to provide. Our own data explains the shortfall from the supply side. The Attribution dimension of the Index actually scores better than most, because the measurement infrastructure, analytics, tag managers, advertising pixels, is genuinely widespread. But measurement infrastructure is not attribution. Collecting the events is common; connecting them into a defensible revenue story is not.

The models most teams fall back on make the problem worse. Single-touch attribution, first click or last click, is easy to compute and wrong in predictable ways, over-crediting whichever channel happens to sit at the edge of the journey. Multi-touch and account-based models are closer to reality but require the connected data layer that most stacks do not have, so they stay on the roadmap rather than in production. The result is that a large share of enterprise marketers run on a mix of last-click reporting and educated guesses, and then defend seven-figure budgets on that basis.

This is why the efficiency gap matters to the board, not only to marketing operations. A martech stack that cannot tie activity to pipeline cannot rank its channels, cannot allocate the next dollar with confidence, and cannot survive a downturn when every line item is questioned. The 22 percent of the marketing budget flowing into martech becomes impossible to govern, not because the tools are bad, but because they were never connected into a system that could answer the one question the CFO always asks. Fixing attribution is rarely a matter of buying an attribution tool; it is a matter of building the data and activation layers underneath it first, which is the same conclusion every chapter of this report keeps reaching.

The high-performance stack

Across the 176 companies we could score with confidence, martech stack maturity is top-heavy at the bottom. But the small group at the top is the most instructive part of the whole dataset, because it shows exactly what separates a stack that produces growth from one that only reports on it.

The distribution is blunt. More than seven in ten companies sit in the Foundational or Developing tiers, scoring at or below the middle of the scale. Roughly one in five reach Advanced, and 5 percent reach Leading. Top-quartile companies score 57 or higher, while the median martech stack scores 44, enough to see, not enough to act. The scarcity of Leading stacks is not a rounding artefact; it is the central finding, and it holds across every way we cut the data.

The defining trait of the top tier is not size. It is balance. Bottom-quartile stacks are lopsided, a pile of measurement and consent tooling with very little behind it. Top-quartile stacks carry every link of the chain: analytics feeding a unified data layer, that data feeding automation, experimentation and conversational engagement, and all of it under real governance. When we compare the two groups dimension by dimension, the gap is widest exactly where it matters most, in Integration, in Revenue Alignment and in the activation-heavy Utilisation proxy. High performers are not the companies that bought the most tools. They are the companies that connected the tools they bought, and then used them.

It is worth being precise about what the top tier is not. It is not the biggest spender, and it is not the company with the longest vendor list on its stack-diagram slide. Several of the highest-scoring companies run comparatively lean stacks in which every tool has a clear job and a clear connection to the next one. Several of the lowest-scoring companies run sprawling stacks that would look impressive in a procurement review and produce very little. Discipline, not spend, is the differentiator, and discipline is learnable in a way that budget is not.

This mirrors the outside evidence closely. Gartner puts martech utilisation at 49 percent and counts only 15 percent of organisations as high performers. Our independent, technology-observed method lands at a 5 percent Leading tier and a median stack that solved measurement and stopped. Two different lenses, pointed at two different samples, describe the same ceiling: the enterprises winning at marketing technology are a small, disciplined minority, and what defines them is repeatable.

The 2027 stack model

The answer to the efficiency gap is not another tool. It is a shape. A martech stack that produces growth carries four layers, in order, each feeding the next, and the data in this report shows that most enterprises are missing the middle two. The value of the model is not the list of categories; it is the sequence and the connections between them.

Each layer only works if the one beneath it is real. Foundation without Activation is measurement theatre, a team that can describe its audience in exhaustive detail and do nothing with the description. Activation without Intelligence is spray-and-pray at scale, a lot of contact with no learning loop to improve it. And none of it is defensible without Governance running through every layer, because a stack that cannot manage consent and data quality is one regulatory change away from a crisis. The companies in the Leading tier are simply the ones whose martech stack has all four layers built and connected in this order.

For most organisations the roadmap is not to add tools at the edges. It is to build the two middle layers the measurement layer was always supposed to feed. That usually means fewer new logos, not more: consolidate the duplicated measurement tools, stand up a single activatable data layer, and only then evaluate the activation and intelligence tools that will sit on top. It is also worth remembering that the foundation itself has to be sound; a content layer built on a slow or rigid platform undermines everything above it, which is why the underlying website and technical foundation belongs in the same conversation as the marketing tools, not in a separate one.

The Unified MarTech Efficiency Index

The Index scores each martech stack from 0 to 100 across seven dimensions, using observable signals as transparent proxies and cross-referencing the market-level dimensions against Gartner and Forrester. Crucially, it rewards balance across the chain rather than raw tool count, which is why a lean, well-connected stack can out-score a sprawling one. The dimension profile tells the whole story of the efficiency gap in a single shape.

Measurement-oriented dimensions, Attribution infrastructure and Data Readiness, score highest. The dimensions that require activation and orchestration, Utilisation and AI Readiness, score lowest. Enterprises are strongest exactly where the tools are cheapest and most default, and weakest exactly where the growth lives. Read from top to bottom, the chart is a ranking of how hard each capability is to buy and deploy, which is not a coincidence; it is the efficiency gap expressed as a scoring profile.

The segment spread is instructive and mostly intuitive. Marketing SaaS companies score highest, which is unsurprising since marketing technology is both their product and their craft; they tend to run their own stacks the way they would advise a customer to. Revenue, HR and CX platforms follow closely, all categories where the go-to-market motion is effectively the business. At the other end, fintech and security companies run the thinnest observable martech stacks, shaped by regulation, gated products that hide most of the experience behind a login, and a justified caution about loading third-party scripts onto sensitive pages.

The most revealing segment is AI. Despite the category’s self-image as the most advanced software being built, AI companies sit squarely mid-pack on martech maturity. Selling intelligence, it turns out, is not the same as operating a mature marketing engine, and the newest companies in the sample have often prioritised product over the unglamorous work of connecting a stack. The Index does not reward reputation; it rewards the observable reality of what a company has built and connected, and on that measure the martech stack of the average AI vendor looks a lot like everyone else’s.

The seven dimensions, defined

The Unified MarTech Efficiency Index scores each martech stack on seven dimensions. Each one is a transparent proxy built from observable signals, so the scoring can be checked rather than taken on trust, and understanding what each dimension measures makes the gaps in the earlier charts easier to act on rather than just read.

Stack Efficiency measures how evenly a company covers the value chain. A martech stack that carries measurement, data, activation and governance in balance scores high; one that piles up tools in a single stage scores low, because breadth without balance is precisely what the efficiency gap is made of. This is the dimension that punishes sprawl and rewards coherence, and it is the reason a lean stack can out-score a large one.

Integration measures the presence of a unification layer, a customer data platform or a mature tag-management and data-routing spine, that lets tools act on each other’s signals instead of sitting in silos. It is one of the lowest-scoring dimensions in the study and, as the integration-tax chapter showed, one of the most expensive to leave unaddressed.

Data Readiness measures whether a company can collect, unify and enrich customer data cleanly enough to act on it. Analytics plus a CDP plus an enrichment or intent layer scores well; analytics on its own does not, because raw events are not the same as an activatable customer view. Companies that have invested in measurement and tracking tooling without a unification layer tend to score high here on collection and low on readiness to act.

AI Readiness measures observable AI-enabled marketing capability, from personalisation to conversational engagement to machine-driven experimentation. It is the weakest dimension across the whole sample, which is why this report treats AI as something a stack earns rather than something it simply buys.

Attribution measures the infrastructure needed to connect marketing activity to outcomes: analytics depth, tag management and advertising measurement. High infrastructure scores do not guarantee good attribution, but their absence guarantees bad attribution, which is why the dimension scores relatively well even though the outcome it enables, provable ROI, remains rare.

Revenue Alignment measures whether the stack is wired to a revenue motion, through marketing automation and an ABM or intent layer that ties marketing to pipeline rather than to vanity metrics. Companies that treat marketing as a demand-capture function score higher than those that treat it as a brand-awareness function.

Utilisation is an activation-ratio proxy: how much of a company’s detected martech stack sits in activation stages versus measurement stages. It is the clearest single signal of the built-to-measure problem this report documents, and it scores among the lowest of the seven dimensions.

The cost of the efficiency gap

The efficiency gap is not an abstraction. It has a price, and the price is large. Marketing technology takes roughly 22 percent of the marketing budget, and Gartner finds that only about 49 percent of that capability is actually used. For a company spending ten million on marketing, that implies well over a million dollars of martech capability sitting unused every year, before counting the staff time spent maintaining tools nobody activates. The unused half of the martech stack is, in cash terms, one of the largest recurring wastes on the marketing budget.

The waste compounds well beyond the licence fees. Siloed data lowers marketing ROI by 15 to 20 percent according to independent analysis, which on the same budget is a far larger number than the software itself. Add the roughly two-thirds of marketers who cannot accurately measure ROI, and a large share of the entire marketing budget, not just the martech line, is being spent without a reliable way to know what it produced. The efficiency gap therefore taxes the whole marketing function, not only the technology within it, and it does so quietly, because none of these costs appear as a single line an executive can point to.

There is an opportunity cost layered on top of the cash cost. Every quarter a company spends adding tools to a disconnected martech stack is a quarter it does not spend building the activation and intelligence layers that would compound. The distance between a Leading stack and a median one is not a few percentage points of efficiency; it is the difference between a marketing engine that learns and improves and one that reports on the same problems month after month. For most enterprises the highest-return marketing investment available is not a new channel or a new tool, it is closing this gap, and a structured marketing consulting engagement or a focused search and content audit is usually the fastest way to find where to start.

Common failure patterns

Across the 190 companies, the low-scoring martech stacks tend to fail in a handful of recognisable ways. Naming the patterns makes them easier to catch in your own organisation before they calcify into a permanent shape.

The measurement trap. The most common pattern by far. A team invests early and heavily in analytics, tag management and dashboards, declares the martech stack mature, and never builds the activation layer that the measurement was supposed to feed. These companies can describe their audience in exhaustive detail and do almost nothing with the description. The fix is not more measurement; it is a customer data platform and an activation layer that finally give the data somewhere to go.

Tool sprawl without a spine. A team buys a credible tool in every category, from AI search tooling to programmatic content tooling, but connects none of them, so the stack looks impressive on a procurement slide and produces very little. The fix is consolidation and integration, not another purchase; often the single highest-return move is to remove tools rather than add them.

The AI-first mistake. A team buys AI on top of a stack that was built only to measure, expecting the intelligence to compensate for the missing activation and data layers. It cannot. The AI inherits the stack’s inability to act and produces the same disconnected outputs faster. The fix is order of operations: connected data and activation first, intelligence on top.

Compliance ahead of activation. A team staffs the consent and privacy layer, which is easy to justify as risk reduction, faster than the growth layer, which requires making a case for upside. The stack ends up well-governed and under-activated, which is safer but not more effective. Governance matters, but it should span an activation layer that exists, not substitute for one that does not.

From Foundational to Leading: the maturity path

The four tiers of the Index are not just labels; they describe a path, and most companies climb it in the same order. Knowing where your martech stack sits today, and what the next tier actually requires, turns the benchmark from a verdict into a roadmap.

Foundational. A stack at this level has solved measurement and little else. Analytics, a tag manager and a content platform are present; a unified data layer, automation and experimentation are not. Thirty-six percent of companies sit here. The work is not to buy more measurement; it is to choose a single activatable data layer and a first activation tool, and to make sure the content foundation underneath is fast and flexible enough to build on.

Developing. A stack at this level has added some activation, usually marketing automation and perhaps an ABM layer, but the tools are not fed by unified data, so each runs on its own partial view of the customer. Another thirty-six percent of companies sit here. The defining move out of this tier is unification: connecting the automation and demand tools to a single customer view so they stop contradicting each other.

Advanced. A stack at this level has built the intelligence layer, a customer data platform and experimentation, and wired attribution toward pipeline. Twenty-three percent of companies reach it. These stacks learn from their own activity, which is the capability that separates a marketing engine from a marketing report. The work here is depth and discipline, running experimentation as a standing practice rather than an occasional project.

Leading. Just five percent of companies reach the top tier, where all four layers are built, connected and governed, and AI sits on top of clean data with a mechanism to act. These stacks are frequently leaner than the sprawling Developing stacks beneath them, because every tool has a job and a connection. Reaching this tier is less about acquisition than about the patient work of turning software into a system.

The jump that matters most is from Developing to Advanced, because that is where the unified data layer gets built and the martech stack stops being a collection of tools and starts being a system. It is also the jump most companies stall on, because it requires connecting things rather than buying them, and connection is harder to put on a slide than a new logo. Teams that get through it usually do so by treating the data layer as a single prioritised project rather than a background task, and by pairing it with the revenue and sales tooling that gives the unified data somewhere to prove its value.

What to do next

The pattern in the data points to a clear sequence of work, and the sequence matters more than any single purchase. It does not start with buying. It starts with connecting what is already owned, then deliberately building the two middle layers the measurement layer was meant to feed. The following ninety-day and twelve-month priorities are ordered so that each step earns the right to the next.

The first ninety days

Begin with an honest inventory of the real martech stack: every tool paid for, mapped to the four-layer model, with actual usage flagged against licence cost. Most teams discover that a meaningful share of the budget is flowing to tools nobody logs into, and reclaiming that spend funds the rest of the work without a new budget request. Next, consolidate the duplicated measurement and consent tools, which are the most common source of sprawl, and redirect the savings toward the missing middle. Then connect the foundation to a single activatable data layer, a customer data platform or a warehouse-native equivalent, before buying anything new; this is the keystone that every later step depends on. Finally, stand up one closed attribution loop from a single campaign to pipeline, even a manual one, so the team establishes a baseline it can improve against.

The next twelve months

With a connected foundation in place, build the Activation layer deliberately: automation and personalisation fed by the unified data, not bolted on beside it. Install an experimentation practice as a discipline rather than a tool, so the martech stack learns from every campaign instead of only reporting on it. Deploy AI only on top of connected data and a working activation layer, where it can compound rather than merely demonstrate. And govern continuously, treating data quality and permissions as a standing function rather than a consent banner at the edge of the site. Teams that need help sequencing this often start with a focused audit of their organic and search foundations, because that is usually where the gap between owned capability and used capability is easiest to see and quickest to close.

The stack rationalisation checklist

• List every martech tool and its annual cost, and flag the ones with no active user in the last quarter.
• Map each tool to one of the four layers; where two tools do the same job in the same layer, consolidate to one.
• Confirm a single, named system of record for customer data; if there is more than one, that is the first thing to fix.
• Trace one real customer journey end to end and mark every point where data has to be moved by hand.
• Redirect the reclaimed budget to the missing middle layers before approving any new tool.

The AI readiness checklist

• Confirm the AI has a unified customer view to reason over, rather than a set of disconnected sources.
• Confirm there is an activation layer for the AI to act through, so its output actually does something.
• Confirm there is an experimentation loop so the AI’s decisions can be measured and improved over time.
• Start with one narrow, measurable use case rather than a broad autonomous agent, and expand only once it works.

Where content is the constraint rather than the connection, the same discipline applies: a thin or generic content layer starves the activation tools above it, which is why the content engine belongs in the rationalisation conversation alongside the tools.

None of this requires a bigger martech stack. In most cases it requires a smaller, better-connected one. The companies that will separate themselves over the next two years are not the ones that buy the most AI or the most tools; they are the ones that do the patient, unglamorous work of turning a pile of software into a system. That work is available to any team willing to start with connection rather than acquisition, and it is the single highest-return project most enterprise marketing organisations have in front of them today.

The two-year outlook

If the enterprise martech stack is built to measure today, the pressure over the next two years will come from the activation side, and it will arrive from several directions at once.

The first is AI, which raises the floor and the ceiling at the same time. Companies that already built connected data and activation layers will compound their advantage, because AI on top of a real system produces real gains. Companies that bolted AI onto a measurement-only martech stack will watch the gap widen, because they are now paying for intelligence they cannot act on. The distance between the Leading tier and everyone else is more likely to grow than to shrink, which makes the connection work more urgent rather than less.

The second is a shift in how buyers research. A growing share of enterprise buying journeys now begins inside AI answer engines rather than a list of blue links, so a brand that is invisible to those engines loses demand at the very first step, before any of its martech stack gets a chance to work. Staying visible there is a distinct discipline, closer to entity and citation building than to classic keyword optimisation, which is why answer-engine optimisation and AI-overview tracking have moved into the activation layer rather than sitting beside it.

The third is budget scrutiny. As marketing budgets face harder questions, the inability to tie activity to pipeline stops being an operational inconvenience and becomes an existential one. Teams that can prove contribution will defend and grow their budgets; teams that cannot will have them cut, regardless of how sophisticated the stack looks on paper. Attribution, and the connected data layer it depends on, moves from a nice-to-have to a survival requirement.

Retention is the quiet fourth pressure. As acquisition costs rise across channels, the value of activating existing-customer data through lifecycle and messaging channels grows, and that activation depends on exactly the unified data and automation layers most stacks are missing. The companies that treat retention as a data-and-activation problem rather than a campaign problem will pull ahead.

The strategic implication is consistent with everything in this report. The winning move over the next two years is not to buy the newest tool but to build the connected system the newest tools quietly assume you already have. That system is what turns AI from a demo into an advantage, turns answer-engine visibility into pipeline, and turns a marketing budget from a cost that has to be defended into an investment that can be proven. The martech stack that wins the next two years is not the largest one; it is the one that was built to convert.

Benchmark your martech stack

Select your category and check the layers your marketing team runs today to see your Unified MarTech Efficiency score against the 190 companies in this study, and exactly where your own chain thins out.

Reading your own results

A single score is a starting point, not a verdict. The more useful output of the benchmark is the biggest-gap flag, because it tells you which layer to build next rather than simply how you compare. A company scoring in the Developing tier with its weakest dimension in Integration should not go shopping for more activation tools; it should connect the ones it already has. A company weak in Attribution should not buy an attribution product; it should build the unified data layer that attribution depends on. The gap, not the number, is the instruction.

The category benchmark matters just as much as the overall one. Comparing a fintech stack against the Marketing SaaS average is unfair to the fintech team, because the two operate under very different regulatory and product constraints; comparing it against other fintech companies is far more actionable. Use the segment averages as the realistic bar for your own sector, and treat the overall average as context rather than a target. A stack that looks weak against the whole sample may be strong for its category, and the reverse is also true.

It is worth checking your own result against reality, because the score measures the observable footprint, not the internal one. If your team runs a mature martech stack that simply does not expose its tools on the public site, the score will understate you, and the honest response is to verify against your own tool inventory rather than accept the number. If the score confirms what you already suspected, that measurement is solved and activation is thin, then it has done its job by pointing at the specific layer to build.

Either way the value is the same: a structured way to see whether your martech stack is built to measure or built to convert, and a clear next step if the answer is the former. Teams that want a second pair of eyes on that next step often run this kind of audit as part of a growth consulting engagement, and we publish more of this research in the Unified Platforms resources library as it is refreshed.

How this was measured

We detected the publicly observable marketing technology on the homepage and pricing pages of 190 enterprise SaaS companies, drawn from the Forbes Cloud 100, the BVP Emerging Cloud Index and a set of public enterprise-SaaS leaders. For each company we recorded the presence of tools across the value chain, then scored each martech stack from 0 to 100 on seven dimensions: Stack Efficiency, Integration, Data Readiness, AI Readiness, Attribution, Revenue Alignment and Utilisation. Each dimension is scored from observable signals as a transparent proxy, and the market-level dimensions are cross-referenced against Gartner and Forrester rather than inferred.

The method has honest limits, and stating them is part of the point. Detection reads the marketing technology that is visible at page load, so the figures are a lower bound: tools injected later through a tag manager, or running only inside logged-in product, are not counted, and 15 companies with insufficient signal were excluded from scoring. The Index therefore describes the detectable martech footprint, not the full internal stack, and the true numbers for advanced-layer adoption are if anything slightly higher than what is shown here. Every Unified figure traces to a logged, re-runnable scan, and every external figure is attributed to its published source, so the whole analysis can be checked rather than taken on trust. The full research hub, including the underlying category breakdowns, is published at the Unified Platforms resources library.

Frequently Asked Questions

What is a good martech stack efficiency score?

In this study the average enterprise martech stack scores 43 out of 100 and the top quartile scores 57 or higher. A score above 55, the Advanced tier, means your stack carries genuine activation and intelligence layers rather than just measurement. Anything below 35 signals a stack that can observe its audience but cannot act on it, which is the most common and most expensive position to be in.

Why do bigger martech stacks not perform better?

Because value in a martech stack comes from balance and integration, not from tool count. A stack loaded with analytics and consent tools but missing a customer data platform, experimentation and automation scores low no matter how many logos it contains, because the pieces that would turn measurement into revenue are absent. Adding more tools without connecting them raises coordination cost faster than output, which is why several of the highest-scoring companies in the study run comparatively lean stacks.

What is the difference between measurement and activation in a martech stack?

Measurement is the set of tools that answer what happened: analytics, tag management, dashboards. Activation is the set of tools that do something about it: automation, personalisation, experimentation, conversational engagement, all fed by a unified customer view. Most enterprises have solved measurement and stalled at activation, which is the core of the efficiency gap this report documents.

How should an enterprise start closing its efficiency gap?

Start by inventorying the real stack and cutting duplicated measurement tools, then connect the foundation to a single activatable data layer before buying anything new. Only after that unified layer exists should you invest in activation, experimentation and AI, in that order. The sequence matters more than the specific vendors, and most of the early wins come from connection and consolidation rather than new purchases.

Does AI fix a weak martech stack?

No. AI amplifies whatever stack it sits on. Deployed on connected data and a working activation layer it can compound results; bolted onto a stack that was built only to measure, it produces the same disconnected outputs faster and more confidently. This is why AI Readiness is the weakest dimension in the Index and why 45 percent of martech leaders report that vendor AI agents underdelivered.

How many tools should an enterprise martech stack have?

There is no correct number, and tool count is a poor proxy for capability. Some of the highest-scoring stacks in this study are comparatively lean, with every tool connected to the next; some of the lowest-scoring are large and sprawling. The right question is not how many tools you have, but whether they cover the value chain in balance and are actually connected to each other.

What is the difference between a CDP and a data warehouse for martech?

Both can serve as the unification layer, and the Index counts either. A customer data platform is purpose-built to collect, unify and activate customer data for marketing; a warehouse-native approach uses the company’s central data warehouse with tools that model and route data back out to marketing systems. What matters for the score is that a single activatable customer view exists, not which architecture delivers it.

Which martech categories are most commonly missing?

Across the 190 companies, the least-adopted categories are conversational tooling at 20 percent, customer data platforms at 28 percent, and experimentation at 31 percent. These are exactly the activation and intelligence capabilities that turn measurement into revenue, which is why their absence defines the efficiency gap and why adding them is usually higher-return than adding yet another measurement tool.