Gartner's 2023 Marketing Technology Survey found that marketers used 33% of their martech stack's capabilities, down from 58% in 2020. A more recent analysis pegged that number at 49%, per Advait Sontakke's reporting on updated Gartner survey findings. Either way, half (or more) of the software your GTM teams touch every day sits idle. That's not a feature gap. It's dead weight on your pipeline.

The instinct when pipeline velocity slows is to buy something new. A predictive scoring tool. A conversation intelligence suite. An AI prospecting layer. Six months and six figures later, reps are toggling between five dashboards and manually logging data into systems that don't talk to each other. The Demand Gen Report calls this the "tool-forward" trap: purchasing software first and forcing operations to fit around it, which adds administrative burden and disconnects systems rather than fixing the original problem.

Revenue Drag Is a Workflow Problem Before It's a Tech Problem

The concept the Demand Gen Report formalized as "pain backward" starts somewhere different. Instead of asking "which tool solves this?" you ask "where exactly does revenue work slow down?" Lost context at a handoff. An SDR re-asking a question the prospect already answered on a form. An AE spending two hours a week updating CRM fields nobody reads. Those are workflow and ownership failures, not software failures.

Mike Ogden, who leads RevOps at Cloudflare, put it plainly in an interview with Terret.ai: tools are meant to reinforce the process, sitting on top of how the business should work. When you skip the process step and go straight to procurement, you automate the broken thing. Chris Stinson made a related point on LinkedIn: adding software without diagnosing the underlying behavior can scale the same failure (inflated pipeline, ignored CRM, inaccurate forecast). Stefan Mersch's framing is the one I'd steal for your next ops review: "Map friction, not features."

The Diagnostic Before the Demo

Before you evaluate a single vendor, trace one lead's journey from first digital signal to closed deal. Write it on a whiteboard. Mark every point where buyer context gets dropped or manually re-entered. That's your context-loss audit.

J. Scerri, writing about RevOps stack problems, argued that many "tooling problems" are actually decision, definition, and ownership problems. You need to understand your funnel, your systems, and your users' pain before rebuilding anything. James McKay made a similar case: a new RevOps leader's first job is to ask "what's broken?" and solve real operational failures, not arrive with a pre-set framework. The pattern across practitioners is consistent. Diagnosis first. Tool second, after you've proven the workflow manually.

Run It This Week: A 30-Day Manual-First Experiment

Pick one handoff that you suspect is leaking context (marketing-to-SDR is usually the richest target). Run the workflow manually for 30 days. No new software. Document the messaging, the data passed, and the buyer progression at each stage. Prove the handoff works cleanly before you automate it.

Success metric: SDR-to-AE conversion rate at the specific handoff point. Guardrails: SDR activity volume doesn't drop more than 10%. Stop-loss: if conversion rate is flat after 30 days with full compliance, the handoff wasn't the bottleneck; move to the next friction point.

Only after the manual process reliably improves deal velocity do you bring in technology, and strictly to remove the manual labor you've now documented.

Integration Is the Constraint, Not Features

A CMSWire report found that 65.7% of organizations named data integration as their biggest stack-management challenge, based on MarTech's "State of Your Stack" survey. That should change how you evaluate vendors. The question isn't "does this tool have the feature I want?" It's "does this tool reduce the number of manual steps between signal and action, and does it integrate cleanly with what I already own?"

For every new point solution you add, sunset an underperforming one. The martech landscape counted 15,384 products in 2025 according to MarketingHubMedia. The supply of tools will never be the constraint. Your team's capacity to adopt and integrate them will.

Pain-backward building is slower at the start. You'll spend weeks mapping workflows and running manual processes before you buy anything. Leadership will ask why you haven't deployed the new platform yet. The trade-off: you lose speed in procurement, but you gain accuracy in diagnosis. And you stop paying for capability your team never activates.

Some stack complexity is justified. Multi-product lines, multiple regions, compliance requirements, different GTM motions: these create real reasons for more tools. The issue isn't "more tools = bad." It's unmanaged overlap and unclear ownership. Replacing tools without fixing definitions and ownership recreates the same problems on a new platform.

The hardest part of this approach isn't the framework. It's the discipline to sit in the discomfort of an unsolved problem long enough to actually understand it, instead of buying your way out of the conversation with a new license.