Gartner puts utilization of the average enterprise martech stack at about 49%. Only roughly 15% of organizations qualify as high performers with positive ROI. That is the uncomfortable backdrop to a market worth $551.96 billion in 2025, projected to exceed $2,380 billion by 2033, with more than 15,000 solutions for buyers to pick from.

The problem, per a CX Today analysis, is not tool quality. It is architecture. Most stacks were assembled to push campaigns out the door, track channel activity, and keep dashboards busy — not to support revenue operations technology, shared customer context, or clean handoffs between marketing, sales, and service. That is why ROI stays hard to prove.

Measured vs. asserted

The strongest data points in the analysis come from third-party research, and they are worth separating from the argument built on top of them:

• McKinsey's 2025 martech work found that none of the roughly 50 senior Fortune 500 marketing leaders interviewed could clearly articulate martech ROI. Its research also showed 34% of martech buyers and decision-makers cite under-skilled talent as a major hurdle.
• Salesforce's 2026 sales research says sales teams use an average of eight tools, 42% of reps report tool overload, and 19% of sales data is inaccessible.
• McKinsey's 2025 State of AI found 88% of organizations use AI in at least one business function, but only about a third have started scaling it, and just 39% report enterprise-level EBIT impact.
• Microsoft's 2025 Work Trend Index found 82% of leaders call this a make-or-break year for rethinking strategy, and 81% expect agents inside their AI strategy within 12 to 18 months.
• BCG found aligning customer-facing teams drives 10%–20% gains in sales productivity, 100%–200% gains in digital marketing ROI, and 30% lower go-to-market expense. Gartner expects 75% of the highest-growth firms to run a RevOps model by the end of this year.

The BCG and Gartner figures are vendor-adjacent consulting claims about outcomes; the utilization and survey numbers are direct measurements. The analysis does not disclose the samples behind the BCG and Gartner projections, so treat them as directional.

Where stacks fail

The analysis identifies four failure patterns, each tied to how the stack was built:

Tool overload. Teams add point solutions one urgent purchase at a time — an AI scoring tool here, a journey tool there, reporting middleware because nobody trusts the dashboard. The result is "maturity debt" across integration, data, process, and vendor dimensions. Bloated stacks slow launches and make every change harder.

Disconnected data. When marketing sees an engaged lead, sales sees a stuck deal, service sees an open ticket, and finance sees an overdue payment, that is four versions of one customer. The proposed fix is "shared customer memory" rather than "integration": a synced record is not a live context layer. The practical test — when something important happens, does the whole system behave differently? A stalled opportunity or renewal risk that doesn't travel breaks routing, suppression, and attribution.

Optimized for activity, not outcomes. Teams measure sends, opens, clicks, and MQL volume while leadership asks about win rate, CAC payback, retention, and forecast confidence. The stack was never designed to connect customer signals to revenue movements finance trusts.

Adoption gaps. A big stack with weak adoption is expensive shelfware with APIs. The problem worsens after procurement, when teams must operationalize governance, data quality, and ownership. AI exposes the gap: tools roll out faster than the work around them gets fixed.

What changes the math

The analysis argues RevOps reframes the buying question from "what can this platform do?" to "what happens to lead quality, routing, conversion, retention, and forecast confidence if this tool enters the system?" A well-wired stack improves routing, suppresses bad outreach, cuts duplicate touches, and keeps service issues from colliding with sales motions.

For CIOs evaluating additions, the article prescribes five pre-deployment checks: architecture fit (KPMG found only 20% of even the strongest marketing-IT relationships involve four or more groups in martech selection, and nearly half assess three or fewer integration dimensions), data quality and identity handling, real-time behavior rather than integration claims, governance and AI transparency, and a first-90-days ownership plan with named owners and role-based training.

The measurement framework stacks three tiers: executive metrics (pipeline, win rate, CAC payback, LTV:CAC, forecast accuracy), funnel metrics (MQL-to-SAL, speed-to-lead, stage conversion, deal velocity), and trust checks (sync latency, duplicate rate, identity-match quality, consent coverage). The author's threshold is blunt: if you cannot follow the path from signal to action to opportunity to revenue, you are still measuring a campaign stack.

With Gartner expecting three-quarters of the highest-growth firms on a RevOps model by year-end, and AI agents entering the stack within 12 to 18 months for most enterprises, the pressure to convert tool accumulation into revenue accountability is likely to intensify before the market doubles again.