The Real Problem Isn’t That PoCs Are Expensive - It’s That You’re Spending the Same Way on Every Deal

Most presales teams don’t have a PoC ROI problem. They have a PoC uniformity problem.

When a deal sitting at 15% win probability gets the same SE hours, the same lab environment, and the same executive briefing as one sitting at 80%, the maths breaks down before the PoC even starts. You can measure ROI all you like, but measurement only helps if you’re measuring the right inputs against the right deal signals. And almost nobody is.

The default behaviour in most SE teams is depressingly predictable: a quota-carrying rep asks for a PoC, and the presales machine spins up. Same process, same effort, same resource commitment. Not because anyone decided this was optimal, but because nobody decided anything at all. The request came in. It felt urgent. Someone started building a custom environment at 9pm on a Tuesday.

This isn’t a motivation problem. Your SEs aren’t lazy or misguided. It’s a prioritisation architecture problem - the absence of a formal system for matching resource intensity to deal quality. Without that system, effort flows toward whoever shouts loudest, not toward whatever’s most winnable. And SEs default to effort-based proof of value: we worked hard, therefore we deserve the win. Which is a lovely sentiment, and also completely uncorrelated with revenue.

Consider an SE team of four supporting 30 active PoCs simultaneously. Under equal treatment, each deal gets 2 - 3 hours of custom configuration. Shared lab environments groan under the load. Discovery calls repeat the same questions because no qualification gate exists to filter them. Everyone is busy. Nobody is effective.

Now consider a team running a tiered model. Tier 1 deals - high ICP fit, champion identified, budget confirmed - get dedicated SE time and custom environments. Tier 2 gets guided self-service with async support. Tier 3 gets a structured “no” with a redirect to documentation or a partner channel.

The second team closes fewer PoCs. They also win more of them. And their SEs don’t burn out by Q3, which turns out to matter quite a lot when you’re trying to retain people who could earn more money almost anywhere else.

What “PoC ROI by Resource Type” Actually Means

PoC ROI by resource type means calculating the return on each category of presales investment separately, then comparing those returns across deal segments. Not total PoC cost - which most teams track loosely, if at all - but the return on SE hours versus infrastructure costs versus executive time versus custom development work.

Without that breakdown, you can’t know whether your custom integrations are actually closing deals or just creating the feeling of momentum. And feelings, while important in many contexts, make for terrible capacity planning.

There are four primary resource types that presales teams actually spend:

SE time - discovery, configuration, enablement, Q&A. The most fungible and most frequently over-committed resource you have.

Infrastructure - lab environments, sandbox licences, cloud compute. Feels free because it’s often amortised, but has real cost and real contention.

Executive and specialist involvement - solution architects, product managers, field CTOs. Scarce, high-signal, and almost never tracked against outcomes.

Custom work - scripts, integrations, tailored demos, one-off configurations. The resource type most likely to be deployed on deals that don’t deserve it, because it feels like the thing that will make the difference.

Almost no team correlates these types to win outcomes. Which means almost no team can answer the question: “Is our custom integration work actually moving win rates, or are we just building bespoke demos for deals that were going to lose anyway?”

This is where a framework helps. Not because frameworks are inherently wonderful - most of them exist primarily to give consultants something to put on a slide - but because this particular problem benefits from a shared vocabulary.

The RISE Model

Resource type. Intensity level. Signal threshold. Exit criteria.

Each PoC decision runs through these four dimensions before resources are committed.

Resource type: Which category of resource is being requested? SE time, infrastructure, executive involvement, custom work - or some combination?

Intensity level: How much? Not “some” or “a lot,” but actual hours, actual cost, actual people.

Signal threshold: What deal signals justify this intensity? ICP fit score, sales stage, champion access, budget confirmation - the specific evidence that this deal warrants this level of investment.

Exit criteria: What does the deal need to demonstrate, and by when, for this resource to keep flowing?

The model gives SEs a vocabulary to push back on low-signal requests without sounding uncooperative. Instead of “I don’t have time for this” - which is both unhelpful and probably untrue, since the problem isn’t time but allocation - it becomes: “This deal hasn’t hit the signal threshold for dedicated SE hours yet. Here’s what needs to be true first.”

That’s a different conversation entirely. One that treats presales capacity as a portfolio to be managed, not a service to be rendered on demand.

How to Calculate ROI When Win Rates Are the Only Output That Matters

Calculate PoC ROI per resource type by tracking: hours or cost invested, multiplied by average deal size, multiplied by win rate for deals using that resource type. Then compare across deal tiers.

Most presales teams resist this because they don’t control CRM data, don’t have a formal cost-per-hour model, or feel - reasonably - that quantifying their work reduces it to a spreadsheet exercise. The goal isn’t to justify every SE’s existence with a number. It’s to identify patterns across resource types and deal outcomes so that future allocation decisions are defensible. Even rough data beats no data.

A worked example, with numbers that are realistic enough to be useful and round enough to be readable:

Resource Type A - SE-led, full custom PoC. 10 deals over one quarter. Average 40 hours each. 35% win rate. Average deal size £120K. Total SE hours: 400. Revenue from wins: £420K. Return per SE hour: £1,050.

Resource Type B - SE-guided, templated PoC. 8 deals. Average 12 hours each. 55% win rate. Average deal size £85K. Total SE hours: 96. Revenue from wins: £374K. Return per SE hour: £3,896.

Resource Type C - Self-service with async SE support. 2 deals. Average 3 hours each. 40% win rate. Average deal size £60K. Total SE hours: 6. Revenue from wins: £48K. Return per SE hour: £8,000.

Type A looks like the premium offering. It gets the most internal prestige. SEs who run full custom PoCs feel like they’re doing the most important work. But the return per SE hour is roughly a quarter of what Type B delivers and an eighth of Type C.

The insight isn’t “stop doing full custom PoCs.” Some deals genuinely require them, and those deals tend to be large enough to justify the investment. The insight is that full custom PoCs should be reserved for deals whose signals warrant them. RISE’s signal threshold determines which resource type is appropriate - not the rep’s urgency level, not the prospect’s enthusiasm, and certainly not the sunk cost of whatever happened in the previous meeting.

Which Deal Signals Should Gate Each Resource Type?

Gate each resource type behind specific deal signals, not just sales stage. Stage is a lagging indicator at best and a fiction at worst - we all know deals that have been “Stage 3” for six months because nobody wants to move them backwards.

Most qualification frameworks - MEDDIC, MEDDPICC, BANT - are designed for AEs. They tell you whether to pursue the deal. They don’t tell you how to resource it. The RISE Model’s signal threshold layer translates deal qualification into resource allocation decisions that SEs can own independently.

This matters for SE autonomy. An SE who can make and defend resource allocation decisions is operating at a strategic level. An SE who builds whatever sales asks for is operating as a very expensive pair of hands.

The mapping looks something like this:

Time-boxing deserves its own mention. A two-week PoC deadline forces the champion to prove internal engagement. If they can’t mobilise their own team for a two-week evaluation, that tells you something important about the deal’s actual status - and the appropriate response is to reassess, not to extend the PoC by another month in the hope that enthusiasm will materialise spontaneously. It almost never does.

What the “Stop” Conversation With Sales Actually Sounds Like

The most effective way to stop burning capacity on low-win deals isn’t a policy document. It’s a repeatable conversation structure.

SEs who can say “here’s what this deal needs to show before I invest another 20 hours” are protecting capacity without damaging the sales relationship. The conversation is about signal gaps, not effort refusal. That distinction matters enormously, because “I’m too busy” invites negotiation, while “this deal hasn’t demonstrated X yet” invites collabouration.

In practice, it sounds like: “I can see why this feels urgent. Let’s look at what we know. We don’t have a confirmed budget or an identified economic buyer. That puts this below the signal threshold for a dedicated PoC environment. What I can do is set them up with a templated trial and async support, and if they engage seriously in the next two weeks, we escalate. If they don’t, we’ve learned something useful without spending 40 hours learning it.”

That’s not a refusal. It’s a resource allocation decision with clear criteria for escalation. The rep knows what “yes” looks like. The SE knows what they’re committing to. The deal gets exactly the investment its signals justify.

The RISE Model doesn’t eliminate difficult conversations. It does make them about evidence rather than feelings, which is a significant improvement over the alternative - where every PoC request is either accepted resentfully or declined awkwardly, and nobody learns anything about why.

Presales capacity is finite. Treating it as infinite doesn’t make it so; it just makes the burnout harder to diagnose. The teams that manage their PoC resources as a portfolio - allocating deliberately, measuring by resource type, gating on signals, enforcing exit criteria - don’t just win more. They win more sustainably. Which, given the current market for experienced SEs, might be the more important outcome.

Further Reading

• Challenger Sales Model Summary & Tips - Covers how structured sales training frameworks, including reframing exercises and role-plays, help reps develop the skills to have more deliberate, controlled conversations - relevant context for SEs building the muscle to push back on low-signal PoC requests.
• Challenger Sales Model Includes Training Reps in Three Behaviours - Gartner’s breakdown of the behaviours that distinguish top-performing sales reps, useful background for understanding why SE autonomy and strategic resource decisions correlate with stronger outcomes.