Proof of value that converts: designing the infrastructure pilot that ends in a decision
A free trial is the buyer’s option to do nothing — and a vendor with network access and no owner. The paid, time-boxed proof of value with criteria the buyer writes is how an infrastructure pilot becomes a contract.
In Your Biggest Competitor Is “No Decision” I set out three forcing functions for enterprise deals, and the first of them was: kill free trials and run paid, time-boxed pilots with written acceptance criteria. This piece is that rule applied to the hardest market I know for it — network operations, where a “trial” means a vendor with credentials to production devices, a security architect who was not consulted, and a NOC manager who wanted it and cannot fund it. The pilot design below is the one I run in The Seven C’s Outreach Program™, and it exists for one purpose: to end in a decision.
- A free trial is the buyer’s option to do nothing, indefinitely, at no cost — and in network operations it is also a vendor with access to the network and no owner on the buyer’s side. A proof of value is the opposite: paid, time-boxed, with success criteria the buyer wrote and a decision at the end.
- Seven rules: paid with the fee credited on signature; thirty to forty-five days on one production domain; eight criteria at most, each with a baseline, a target, a method and an owner, signed before day one; the Economic Buyer named in the plan; read-only first; vendor onboarding in parallel; and one of three decisions at the readout — buy, extend for one named gap, or stop.
- Lead the readout with the misses. A vendor that reports what it did not prove is believed about what it did, and the extension-for-one-gap is where that honesty turns into a contract.
1. Why free trials are where infrastructure deals go to die
A free trial sounds generous and costs the buyer nothing, which is precisely the problem. Nothing moves. No budget is created, no owner is named, no date exists. In The Seven C’s the force that closes a deal is Commitment — a decision with a timeline, ownership and next steps — and a free trial is designed to avoid every one of those. In network operations it adds a specific hazard: a vendor holding device access with nobody on the buyer’s side accountable for it. The guardian discovers this in week six, and the deal pauses for a quarter.
The paid, time-boxed proof of value fixes all of it at once. Money moving creates a budget, an owner and a deadline. Written criteria turn “it worked” into a measurement. A named Economic Buyer in the plan means the person who can sign has heard the story before day one. And a decision at the end — one of three, agreed in advance — makes “let’s stay in touch” impossible to say.
2. The seven rules
A fee, credited in full to year one on signature
The fee is material enough to need an owner and small enough for the Champion to defend without an approval cycle. Credited in full if the order form is signed within thirty days of the readout. That is not a pricing decision; it is a qualification mechanism.
Thirty to forty-five days on one production domain
One site, one region, one business unit, one campus — big enough to produce real incidents in the window, small enough to deploy in week one. Open-ended pilots are where momentum dies.
Eight measures at most; four is better
Each with a baseline, a target, a method and a buyer-side owner, in the buyer’s terms. The buyer writes the first column; the vendor proposes the rest; both sign.
Named, engaged before day one, present at the readout
No Economic Buyer, no pilot. If the person who signs has not heard the story first-hand, the next step is a twenty-minute executive conversation, not a proposal.
Access expands only as the buyer approves it
Scoped, revocable credentials, brokered from the privileged-access tool the buyer runs where they have one; an execution record for anything the product does. The security packet goes to the guardian in week one.
The readout is not the first time legal sees the paper
Security review opened in week one; procurement opens the order form in week three; the standard packet, data-flow diagram and subprocessor list are ready before the pilot is proposed.
Buy, extend for one named gap with a new date, or stop
No fourth option, and the buyer agrees to that before day one. A second extension is a stop.
3. Scope: one door, fully
Network operations products tend to open three commercial doors — incidents, changes, resilience. A pilot opens one fully and may touch a second. Three doors in forty-five days is a product tour, not a proof.
| Door | What the pilot proves | What the buyer provides |
|---|---|---|
| Incidents | Past incidents replayed with the product not told the cause; live incidents in the window explained with the evidence attached and written to the ticket | Five to ten past incident tickets with their post-incident reviews, held back from the vendor until after the replay |
| Changes | One real change window compared before and after, with the outcome and the evidence recorded in the change ticket | A scheduled change in week three or four; access to the change record |
| Resilience | Posture and resilience findings — standby pairs out of sync, configuration drift, insecure management settings, contextual exposure — each with its evidence and fix | Nothing beyond access; the buyer validates a sample of findings |
The scope template names the domain, the estate (how many managed elements, which vendors), the integrations (the ticketing system, one telemetry source, configuration access, identity), the deployment model as available today, the access model, what leaves the environment, the buyer’s effort (about four hours across forty-five days, stated so the Champion can defend it), the vendor’s effort, the fee and the credit, and the dates. Day zero is the signed plan; day one is the day access works; the readout is on the calendar before either.
4. Success criteria the buyer would defend to their boss
Each criterion has five parts, and the buyer writes the first. Here is the shape for a network assurance pilot — examples to adapt, not targets to copy:
| Measure | Baseline | Target (example) | Method | Owner |
|---|---|---|---|---|
| Past incidents explained correctly without being told the cause | Not measured | [N] of [M] replayed incidents, as judged against the post-incident review | Week-two replay, blind to the review, scored by two engineers | Director of network operations |
| Time from alert to explanation on live incidents | [hours, from ticket history] | Median under [X] minutes | Ticket timestamps versus product output, all live incidents in the window | NOC manager |
| Change window verified before and after, outcome in the ticket | Manual, [hours] per window | One real window verified; variance against intent reported | Week-four change; vendor and buyer compare | Change manager |
| Resilience findings validated | Not measured | [N] findings; a sample of [S] validated true | Week-three report; buyer validates the sample | Network architect |
| Senior-engineer hours per incident | [estimate] | Reduced by [X]% on explained incidents | Engineer log in the window | Director of network operations |
| Security review | — | Packet accepted, or open items listed with owners by week four | Vendor-risk process | Security architect |
Four rules for the criteria. One honest baseline — “not measured today” is acceptable and often the most persuasive. Targets the buyer would defend upward; the vendor does not inflate and the buyer does not sandbag. A method both sides can run, so no result depends on the vendor’s interpretation. And misses reported: the readout shows every criterion, including the ones missed, because the miss is what earns the extension or the stop.
5. Forty-five days, week by week
| Week | Buyer | Vendor | Output |
|---|---|---|---|
| 0 | Signs the plan and criteria; names the Economic Buyer and the readout date; provides the access list | Delivers the security packet; schedules the readout | Signed plan; readout on the calendar |
| 1 | Grants read-only access; identifies past incident tickets (held from the vendor); introduces the ticketing owner | Deploys; connects integrations; discovers the domain | Discovery report: what was found, what was not reachable |
| 2 | Two engineers score the replay | Replays five to ten past incidents, not told the cause | Replay scorecard |
| 3 | Validates a sample of findings; procurement opens the order form | Posture and resilience findings with evidence and fixes; drift report | Findings report |
| 4 | Runs the scheduled change; the change manager compares | Captures before and after; records outcome and evidence in the ticket | Change verification record; security review status |
| 5 | Check-in; engineer-hours log | Live incidents explained and written to the ticket; measures compiled | Interim scorecard |
| 6 | Readout with the Economic Buyer; decision | Readout; decision memo within 24 hours | Buy, extend for one named gap, or stop |
Two thirty-minute check-ins, at the end of weeks two and four, with the Champion and the vendor’s technical lead. No new scope is added at a check-in; a gap is noted for the readout. If the network is quiet in the window, the replay of past incidents and the scheduled change window carry the proof — which is why both are in the plan rather than hoped for.
6. The readout and the decision memo
Forty-five minutes. The plan as signed; the scorecard, criterion by criterion, misses first; three things the buyer’s own engineers said; the commercial summary — the order form already with procurement, the credit, the term options; and the decision. The Economic Buyer is in the room or the readout is rescheduled. The buyer’s engineers present the replay score, not the vendor. And the decision is asked for in the meeting: “we will discuss internally” is answered with “of course — which of the three, and by when?”
The decision memo follows within a day, on one page: the decision requested; what was proven, with results; what was not, with the reason and what would change it; what the team said; the commercial path and the date the credit expires; the first thirty days under each option; and three signatures. It is the document the Economic Buyer forwards, which is why it is one page and why the misses are on it.
7. Commercial design that survives procurement
Commercial Alignment is where good deals die quietly, so the pilot is designed for it. Price by the size of the managed environment, the capability tier and the deployment model — never per engineer seat, never metered per query — because that is how infrastructure teams budget, and the price should read like their budget. A band in conversation after scoping; a number only in the proposal; never in the first email. Standard paper that fits the buyer’s vendor-onboarding process without negotiation. Term options that let the buyer choose their risk: one year with the conversion credit, or multi-year with a price hold. And an extension that means one named gap, one new date, no new fee unless scope changes materially.
8. How the pilot fails, and what prevents it
| Risk | What it looks like | Mitigation |
|---|---|---|
| Access delayed | Week one passes without credentials | Access list in the signed plan; day one is the day access works; the readout moves only by written agreement |
| Scope creep | “While you are here…” | No new scope after day zero; gaps go to the readout |
| Criteria drift | The buyer’s team redefines a pass at the readout | Signed criteria with methods; misses reported honestly, which removes the incentive to argue |
| Economic Buyer absent | The readout happens with the Champion only | Rescheduled; the plan named the Economic Buyer and the date on day zero |
| Security stall | The packet sits in a queue | Delivered in week one; the vendor answers inside five business days; open items with owners by week four |
| Vendor over-claims | A design target quoted as a result | The claims register; every result labelled by method |
The trade-off is explicit. This design proposes fewer pilots than a free-trial motion and starts more that convert, because the ones that start have a budget, an owner, a date and a person who can sign. The companies that run it end each quarter with a register of measured results; the companies that run free trials end it with a dashboard and a forecast made of hope.
The Seven C’s Outreach Program™
The proof-of-value kit — scope template, success criteria, the forty-five-day plan, the readout, the decision memo and the commercial mechanics — comes with the program.
See the program »Your Biggest Competitor Is “No Decision”
The three forcing functions — paid pilots, the Economic Buyer in thirty days, the signed decision plan — that this piece applies to infrastructure.
Read the program »Questions this piece answers
What is the difference between a proof of value and a free trial?
A free trial is open-ended, unpaid and undefined: the buyer can run it forever and nobody owns the decision. A proof of value is paid (with the fee credited to year one on signature), time-boxed to thirty to forty-five days, scoped to one production domain, governed by written success criteria the buyer set, and ends in one of three decisions: buy, extend for one named gap, or stop.
How long should an infrastructure pilot run?
Thirty to forty-five days on one production domain. Long enough to replay the buyer’s past incidents, produce posture and resilience findings, verify one real change window and explain the live incidents that occur; short enough that the readout date is on the calendar before day one and the Economic Buyer’s attention survives it.
What success criteria should a network assurance pilot have?
At most eight, written by the buyer in their own terms, each with a baseline (an honest “not measured today” is allowed), a target the buyer would defend to their boss, a method both sides can run, and a buyer-side owner who validates the result. Typical measures: past incidents explained without being told the cause; time from alert to explanation on live incidents; change windows verified with the evidence in the ticket; resilience findings validated; senior-engineer hours per incident; security review accepted; explanations written to the ticket in the buyer’s format.
Part of the outreach series for companies selling into AIOps and network operations: Signal-Based Outbound for Network Operations, Selling to the NOC, Close the Loop, Don’t Break Up, Infrastructure Buyers Have Heard “AIOps” Before, this piece and Operator Edition No. 05: The Outbound Agent Mandate.
Discovery and qualification
One of the five stages of the Revenue Foundation. Every piece for this stage, in order, is on the Field Guide.
Open »The Seven C’s Outreach Program™
The proof of value is the program’s closing instrument. Signals, the committee, the message spine and the deal-force check get you there.
See the program »New essays and operator tools, as they publish.
A few times a month, by email. No list is sold or shared; one click to leave.