--- title: "How we closed $8,309,795 of commissioning and critical-facilities work for a data center commissioning firm in nine months. | Stone Haven Capital Group" description: "13 programmes at an average of $639,215 each. Nine months. A data center commissioning and critical facilities firm. No new business development hires, no trade association circuit, no waiting for a general contractor to already have the name on a bid list." canonical: "https://stonehaven.capital/showcase/data-center-commissioning" last-updated: "2026-08-22" --- > 13 programmes at an average of $639,215 each. Nine months. A data center commissioning and critical facilities firm. No new business development hires, no trade association circuit, no waiting for a general contractor to already have the name on a bid list. Stone Haven Capital Group Schedule a Call Case study · Data Center Commissioning # How we closed $8,309,795 of commissioning and critical-facilities work for a data center commissioning firm in nine months. 13 programmes at an average of $639,215 each. Nine months. A data center commissioning and critical facilities firm. No new business development hires, no trade association circuit, no waiting for a general contractor to already have the name on a bid list. - $8,309,795 generated in 9 months, from 13 commissioning programmes at an average of $639,215. - 518,070 emails to 103,614 companies, producing 1,036 replies at 1.0%. - 197 interested conversations became 57 held meetings and 13 signed programmes. - What did most of the work: we kept cold volume off the firm's own domain. - What went wrong: we aimed the list at the operator when the contractor held the commissioning scope. 01 ## Did they get a good result? Here it is with the arithmetic showing. In the nine months before we started, comparable new work came to $2,596,811 and arrived through repeat operators and referrals from design engineers. Nine months later the firm had closed 13 commissioning programmes worth $8,309,795, a 3.2x increase, with $26,904,000 of qualified work sitting behind it. What happenedThe number Commissioning programmes closed13 Value of work closed$8,309,795 Average programme value$639,215 Qualified opportunities behind it57, worth $26,904,000 Interested-stage conversations197, worth $62,252,000 Emails sent across nine months518,070 Contacts reached103,614 Direct replies1,036 (1.0%) Positive replies197 (19.0% of direct replies) Bounce rate1.2% (6,217 bounces) Nine months of sending against construction and critical-facilities contacts, as the platform logged it. 02 ## Did you get them in front of the right people? This is the only line in the document worth pushing on. 518,070 emails prove nothing by themselves, because a data center becomes public knowledge long after the commissioning agent has been appointed, and most of the market that is easy to find is already committed. 13 programmes reached contract carrying $8,309,795, an average of $639,215 each. Behind them sit 57 qualified opportunities worth $26,904,000 and 197 earlier conversations worth $62,252,000. Targeting shows up in the conversion rather than in the send count. 1,036 direct replies produced 197 positive ones, 57 of those reached qualified stage and 13 closed. Close to one qualified opportunity in four became signed work, which is what happens when the conversation starts before the scope has been written. Commissioning programmes moving through the client's pipeline, stage by stage. 03 ## What worked, and why ### We kept cold volume off the firm's own domain. The firm's domain carries drawings, test scripts, commissioning reports and invoices for buildings that are already running, and none of that was going near cold sending. We stood up 87 dedicated sending domains carrying 262 mailboxes, each authenticated on SPF, DKIM, DMARC and MX before anything left. Warmup ran on all 262 for the life of the programme, and only part of the estate carried campaign volume at any one time. Whatever was held back stayed idle and rotated in as the domains already running gained age. Each inbox was capped at 20 sends a day. Across nine months the estate carried 518,070 sends and finished at a 1.2% bounce rate, 6,217 bounces in total, with inbox placement holding at 93.0%. Monitoring logged 0 errors and 6 alerts over the whole run. Construction contact data moves quickly, because people follow projects between firms, so holding a bounce rate in that range takes continuous verification rather than one clean build at the start. The 87 sending domains and the 262 mailboxes on them, none of it on the firm's own domain. The domain that talks to operators already under contract never carried a cold email. Every risk in the programme sat on infrastructure that was built to be discarded. ### We built the list around projects still at design stage and cut everything already awarded. The people who decide who commissions a building are directors of construction, VPs of critical facilities and heads of data center operations at colocation and hyperscale operators, together with the preconstruction and MEP leads at the general contractors who build for them. That profile framed the list across the regions the firm can actually staff. The exclusions did as much work as the inclusions. Projects with a commissioning agent already named came out, so did every region the firm cannot put a team into, and so did enterprise server rooms below the size at which integrated systems testing is a separate scope with its own budget line. Between them those cuts removed most of the original pool before any copy existed. A building under construction is a decision somebody already made. Design stage is where the decision still lives. Firmographics tell you who builds data centers. Permit filings, utility interconnection queues, land assembly and design-firm appointments tell you who is about to. ### We ran five touches on widening gaps and moved every positive reply the same day. Five steps, with the wait after each one longer than the wait before it, three variations on the opener and two on each follow-up, all A/B tested. The widening is deliberate against a buyer whose decision may sit a year out. A sequence that finishes inside a fortnight has left the inbox before the question it was written for has been asked. Counting automatic responses, 6,009 replies came back across the programme, 4,973 of them out of office at a 4.8% rate. 1,036 were direct. 197 of those were genuinely positive, and every one went to somebody who could talk about scope, schedule and staffing in the first conversation. The five steps and the waits between them, as the sequence was configured. A commissioning agent is chosen once per building. The purpose of the widening gaps is to still be present when a project moves from concept to design, which is a different job from extracting an answer while nothing has yet been decided. ### The first email read like a note somebody had forwarded on. No greeting and no sign-off. One line of context saying why that specific project was on the sender's radar, then one question and nothing after it. Every send went out as plain text, spun across every line, so no two messages leaving the estate were identical. A director of construction reads forwarded fragments all day and answers them without ceremony. A formal introduction announces a vendor before the question has arrived, and the question is the only part of the email with any chance of being answered. ### We judged the angles on the conversations they started, never on the replies they collected. Angles were judged on whether the reply came from a project still at design stage rather than on how many replies they produced. At a 1.0% reply rate, an angle pulling responses out of finished buildings is worse than an angle pulling none, because it spends a contact that cannot be approached again for a year. 04 ## The four angles we test, in every market Which message a market will answer cannot be worked out in advance, and anyone offering a prediction is offering a guess with confidence attached. The market decides, so we ask it directly. The opener ships four variations at once, one per angle, and reply quality names the winner. These four travel across sectors because they are built on how a decision gets made rather than on what is being sold. 05 ## 1. The named problem State the constraint they already live with, in the words they would use, then ask whether it holds on their side. It costs nothing to answer and the answer tells you where the project is. Most operators we talk to are still choosing a commissioning agent after the design is locked, which is where the retest cost comes from. Is that how it runs on your side? 06 ## 2. The teardown Offer one specific read on something of theirs that is already on the record. Expensive to produce and close to impossible to ignore, so it goes only to the projects worth the hour. We pulled the interconnection filing on the north site and marked two places where the energisation date and the fit-out schedule disagree. Want it? 07 ## 3. Partnership or collaboration Frame it as two parties looking at the same project rather than as a sale. Carries markets where the delivery team is assembled from several firms as a matter of routine. We are looking at the same site from the commissioning side. Worth comparing notes before either of us is committed to a schedule? 08 ## 4. Social proof Put a comparable project in front of them with a real outcome attached. No conclusion has to be drawn for them, and the comparison does the arguing. A colocation operator building at about your scale brought us in at design stage last year and took two weeks out of integrated systems testing. Happy to send what that looked like. 09 ## How we use them All four go out together across a divided list. Within a few weeks the reply data has named the one or two this market wants, and the rest come off before they spend any more of a pool that took months to assemble. Which of the four wins is not forecastable. That one of them will win is. Running four at once is what turns that into arithmetic. The same four port to LinkedIn on shorter wording with the underlying logic unchanged. 10 ## Why this works for data center commissioning specifically The commissioning firms that call us are usually in the same position. The technical depth is real, the operators who have used them once use them again, and every new programme still arrives through a design engineer or a contractor who already knows the principals. That builds a solid business with no forward visibility, and it never reaches an operator putting its first campus into a region where nobody has heard the name. Three features of this market do most of the work: - The building announces itself years early. Land assembly, utility interconnection applications, permit filings and design-firm appointments are all public, and all of them happen well before anybody is appointed to commission the site. That timing signal is searchable, which is the whole reason the channel works here. - One programme pays for the year. At an average of $639,215 with ongoing critical-facilities work behind it, the commercial argument is settled on the first result. Very few channels carry that ratio. - The scope is never optional. Integrated systems testing happens on every critical facility, by somebody, on every project. Nobody has to be persuaded the category matters, which removes the expensive half of the conversation and leaves only the question of who does it and when it is decided. 11 ## What did not work, and what we did about it None of the case above was obvious while the programme was running. It is what nine months taught us, and roughly three of those months paid the tuition. Here is what went wrong, in the order we found it. ### We aimed the list at the operator when the contractor held the commissioning scope. The opening list went almost entirely at colocation and hyperscale operators, on the assumption that whoever owns the building appoints whoever commissions it. The replies said otherwise. Enough of the early positive responses came back as a name at the general contractor that we stopped and audited the pipeline against the contracts, and on a clear majority of live projects the commissioning scope sat inside the construction contract rather than with the owner. We had spent six weeks writing to people who would have had to ask somebody else for permission to hire us. The first pull: every construction and critical-facilities title across the target regions. What we changed: the contractor side went onto the list as a primary target with its own angle written for preconstruction and MEP leads, and owner-side contacts were kept only where the operator self-performs commissioning procurement. The design-stage signals and the hard exclusions then took the pool from 321,562 down to 119,097. The same market once the design-stage signals and the exclusions were applied. One conversation with a construction director, opened eighteen months before the build. ### The list was built out of press releases, so half of it was already too late. Project announcements are the most available signal in this market and among the worst. A release goes out when the site is close to breaking ground, and the commissioning agent was usually appointed twelve to eighteen months before that. Nothing was wrong with the contacts. They were real people, at real projects, with real budgets, and the list quality was fine. We had simply arrived after the only decision we were there to influence, on close to half the projects in the file. What we changed: sourcing moved upstream to permit filings, utility interconnection queues, land assembly and design-firm appointments, and any project already carrying an announced construction start came out unless a later phase was visible behind it. Same market, read eighteen months earlier. ### Replies landed with a regional manager who could not act on them. Positive replies went back to whoever the sequence had addressed, and in a number of accounts that was a regional facilities manager. Bringing in a new commissioning agent is not a decision that seat holds. Rather than say so, several of them held the conversation politely for weeks while nothing moved internally, and three were still sitting open on the day the projects awarded. What we changed: any reply from below the level that can appoint an agent was answered with a short note asking who owns the commissioning scope on that particular project, and the conversation either moved within two days or it stopped. On a programme carrying 13 closings, each of the 197 positive replies is worth protecting one at a time. Thirteen closings leave no room for a mistake to be absorbed quietly. We would rather hand you the three corrections than publish a document that reads as though the first month ran perfectly, because the corrections are the part that carries over to your engagement. 12 ## Why would this work for your business? There are markets this does nothing for. The table is the fastest way to work out which side of the line you are on. This works ifThis does not work if A single won programme is worth six figures to you.The work is priced per visit and repeats at low value. You are appointed long before the work starts, and the signs that a project is coming are public.You are picked at the last minute from whoever happens to be available. You can name the specific projects you want, and there are hundreds of them rather than millions.Your buyer is anybody who owns a building. You can staff a new region if the work lands there.Every new client has to sit inside an hour of your office. A conversation with a stranger can end in a scope document.Nothing moves without an existing relationship or a procurement panel you are not on. The left column describes a method rather than this client. Nothing about the firm was unusual: genuine technical depth, a full order book from people who already knew them, and no route to an operator that did not. What produced the result was process, and process is portable. 13 ## One more thing worth understanding The awkward fact about this market is that the work is bought roughly eighteen months before it is done, and most outbound aimed at it is aimed at the wrong moment entirely. By the time a data center is announced the design is largely fixed and the commissioning agent has usually been named. The public event that looks like the beginning of the opportunity is closer to the end of it. Everything worth reaching sits earlier, in filings, queue positions and hiring that nobody writes an article about. That changes what the channel is for. The first email exists to be sitting in the inbox on the day the question finally comes up, which may be nine months later. The cadence widens for that reason, every follow-up has to still read sensibly a year on, and a good conversation is measured by how early the project is rather than by how warm the reply sounded. It changes what patience means commercially as well. A programme aimed at design stage looks slow through the first quarter and then produces closings that were seeded in month one. Judged on closed work in the first eight weeks it would have been switched off. Judged on how many early-stage projects had entered a conversation, it was moving from the third week, and that was the honest read. 14 ## Before and after Before After New work booked in nine months$2,596,811 $8,309,795 How programmes arrivedReferrals from design engineers and repeat operators 13 closings at an average of $639,215 When a project was first touchedAfter the build had been announced At design stage, before the scope was let Reachable buyers identifiedThe operators the principals already knew 119,097 filtered contacts across operators and contractors Who carried business developmentThe principals, between projects 87 sending domains and 262 mailboxes running continuously 15 ## If you want to know whether your market has this in it One call is enough. Tell us which projects you would want to be commissioning and what one is worth across the build and the maintenance years behind it. We come back with how many of those projects are reachable while the design is still open, the conversation volume that is realistic against that pool, and a straight answer on whether this channel is the right instrument for you. If that answer is no, it arrives on the call. Neither of us gets anything out of nine months spent proving a market was never there. ## Want to know whether your market has this in it? The first conversation is short. You tell us who your buyers are and what one is worth to you. We tell you how many we can actually reach, what the meeting volume looks like, and whether outbound is the right lever for you at all. If we think it is not, we will say so. Book a consultation call → ## Read another one [Defense Technology Supplier](https://stonehaven.capital/showcase/defense-technology-supplier/)[How we built a $38M tracked pipeline for a defense technology supplier in six months.](https://stonehaven.capital/showcase/defense-technology-supplier/)[Employee Benefits Brokerage](https://stonehaven.capital/showcase/employee-benefits-brokerage/)[How we put $6,778,510 of annual commission on the book for an employee benefits brokerage in eight months.](https://stonehaven.capital/showcase/employee-benefits-brokerage/)[All case studies](https://stonehaven.capital/case-studies/)