The Job Description Is the SOW
§1. The open marketing requisition
There is a marketing requisition open at one of your portfolio companies right now. It has been open for somewhere between sixty and a hundred and twenty days. The hiring manager has lost one candidate to a counter offer, interviewed two more who will not commit at the band, and the head of HR has been quietly told the search is moving slowly because the brief is broad. The board update due next week needs a number for marketing capacity and the honest answer is that the seat is still empty.
Look at the requisition itself. The responsibilities section is a list of the work the company expects to get out of the role. The salary band is a pre-approved budget envelope, with benefits and on-costs already loaded in. The level is the seniority gate, signed off by Finance and HR. The hiring manager is named, with authority to direct the work, and the portco CEO has cleared the offer letter the moment a candidate accepts. Every approval the trial mechanic needs is already in place. The req is not a job posting waiting for a candidate. The req is a Statement of Work, written, with a budget envelope already cleared by everyone who has to clear it.
§2. The 35 percent constraint
The FTI 2026 Private Equity AI Radar, published 19 May 2026 from a survey of 200 fund and operating leaders, reports that ninety five percent of funds say AI initiatives are meeting or exceeding their original business case criteria. The same report names the constraint on getting to the next dollar of that case: talent remains the primary constraint to scaling adoption, cited by thirty five percent of respondents. It is the largest of the named constraints in the report.
The talent constraint is structural. The roles a portco needs to run AI-augmented marketing and operations have a six to nine month time-to-productivity even when the seat fills. The candidate pool is thin because the work itself is new, and what passes for evidence on a resume is a list of tools the candidate has used rather than a record of outcomes the candidate has shipped under governance. The fund modelled a marketing capacity ramp that assumed the seats would fill in ninety days. The seats are taking nine months and the productive months at the back of that are eaten by the next reorg.
This is what the FTI line actually says when you read it as a value-creation problem rather than a headcount problem. The fund has the budget. The fund has the mandate. The fund cannot get to the work because the only mechanism it has used to get to the work, namely hiring a human into the seat, has a structural lag the model did not assume.
§3. The trial mechanic
The req is a Statement of Work with a budget envelope and pre-approved sponsors. Use it.
Do not fill the requisition this week. Starting Monday, the JieGou operations team staffs that seat. The first two weeks are at no cost. The hiring manager judges the work against the responsibilities the JD already lists. The requisition stays open the whole time, and at any point in the trial the fund or the portco can pull the trigger on a human hire and walk away from JieGou with no obligation. If at the end of two weeks the work clears the bar the JD set, the engagement converts to the published price at the published cadence. If it does not, the req goes back into the pipeline and the fund has lost two weeks of cost-free trial.
The mechanic is the inversion of the standard procurement cycle. The standard cycle gates the trial behind a procurement process, a master agreement, a security review and an SOW draft, all of which add four to eight weeks before the first measurable work happens. The trial then runs against a custom scope written specifically for it, with no comparable baseline. The buyer is asked to evaluate the trial in isolation, against criteria they had no prior reason to commit to. The JD-as-SOW mechanic skips that whole sequence. The scope already exists, the budget envelope is already cleared, the sponsor is already named, and the evaluation rubric is the document the hiring manager has been using for sixty days. The trial does not need a custom contract because it borrows the contract Finance, HR and the hiring manager already agreed.
§4. Why this is the only structurally fast path
Every other path the fund has tried slips the same way. Backfilling the seat with an agency runs at a multiple of the band, takes thirty days to spin up, and produces deliverables that do not match the responsibilities of the role because the agency sells what it sells rather than what the JD asks for. Standing up an in-house AI capability inside the portco requires hiring against an even longer time-to-productivity at a more senior band, with no pipeline of people who have shipped against an outcome under governance. Outsourcing to a Big-4 transformation programme moves at the wrong altitude. The transformation programme designs the operating model the portfolio should run. It does not staff Monday’s marketing work.
The JD-as-SOW path is the only one that closes the gap between the moment the requisition opened and the moment the work starts. It runs at the band the fund has already approved. It produces deliverables that match the JD by construction. It generates an honest evaluation in two weeks against a rubric the hiring manager wrote themselves. It leaves the human-hire path open the whole time, so the trial costs the fund nothing if the bar is not cleared.
§5. The Big-4 question
“We already work with KPMG.” Some version of this is the most common response when an operating partner first hears the offer. The May 2026 KPMG×Anthropic global alliance was widely placed at PE funds, with the brief framed around portfolio transformation, tax efficiency and the design of AI operating models across the holdings. A fund that has commissioned a programme with a Big-4 partner has a reasonable question about where the JieGou engagement sits next to that work.
The honest answer is that the two engagements do not occupy the same square. The Big-4 partner sells the fund a transformation programme: design of the operating model, sequencing of capability roll-outs across the portfolio, change management, and a multi-year service contract that scales with portfolio size. The engagement is priced at fund level and sits at the strategy altitude. The deliverable is a programme.
The JieGou engagement is the per-portco execution layer the consultants bill the fund to find. The Big-4 programme says marketing operations needs to be governed and consolidated across the portfolio. The JieGou trial staffs the seat at the first portco on Monday and produces the first month of governed work against the JD. The two engagements complement each other. The Big-4 programme is upstream and tells the fund what to do. The JieGou trial is the work itself, in the building, against the requisition that is already open.
The cost-of-time comparison is unflattering for any path that asks the fund to wait. A multi-million-dollar transformation SOW that produces the first piece of execution work twelve months in is paying the consultants to discover, design and recommend. A seventy five thousand dollar trial that staffs Monday and produces the first month of work against the JD is paying for the work to happen. A fund that has already signed the transformation SOW is the strongest candidate for the JD-as-SOW trial, not the weakest, because the trial gives the transformation programme an evidence base it does not otherwise have.
§6. What the first two weeks produce
A typical mid-market portco marketing requisition lists between eight and fourteen responsibilities. Across the open reqs JieGou has read this cycle the recurring items are: own the editorial calendar across owned channels, run the weekly content cadence end to end, manage the CRM-segmented lifecycle programme, run paid social against a defined acquisition target, govern the brand voice across locations or franchisees, instrument and report on funnel performance to the leadership team, and own the agency or vendor relationships that support the function. Each of those responsibilities is a deliverable the JieGou ops team can produce in week one or week two against the JD.
The two-week artifact list, against a typical marketing-ops JD: an editorial calendar populated for the next four weeks, one piece of content shipped per channel per week behind a named-approver gate, a lifecycle programme audit with one segment activation in flight, a paid social plan with weekly creative and the first week’s spend run against the defined target, a brand-voice governance document keyed to the portco’s specific compliance and franchisee constraints, a funnel report with the metrics the hiring manager said they wanted, and a written hand-off on every vendor and agency relationship the role inherits. The evaluation is straightforward: against the responsibilities the JD listed, what shipped against governance in two weeks, and what is in flight that did not. The hiring manager grades the trial. The fund and the portco decide.
§7. The two-trial cap
JieGou is running this trial mechanic at a deliberate two-engagement concurrent capacity through Q3 2026. The capacity is honest. The work behind the trial is staffed by a small operations team running governed recipes with named-approver gates across a thirteen-channel chassis. Adding a third concurrent trial without the capacity behind it would dilute the work on the two trials that are already underway, which is the failure mode the trial mechanic is specifically designed to avoid. An operating partner reading this and recognising the seat is the one open at one of their portcos has a useful window of the next several weeks to take the trial slot. After that, the queue forms.
This is not artificial scarcity. It is the operator-honest answer to the FTI talent number. The seats are scarce because the work is hard, the work is hard because doing it under governance has not been productised yet at the per-portco execution layer, and the productisation is what JieGou is doing one trial at a time.
§8. Close
Every portfolio marketing requisition that has been open more than ninety days is a unit of evidence that the AI category has not yet earned the right to be the answer. The funds that prove the category, on the timeline the talent constraint actually permits, will be the ones whose value-creation teams stopped paying transformation consultants to design the operating model and started running execution against the requisitions the operating model implies. The fastest path to the first dollar of the AI value-creation thesis is the one that uses the requisition you already have, because the requisition you already have is the only Statement of Work the fund has both written and signed.
JieGou is taking two more trials this quarter, run against the open requisition, at no cost for the first two weeks, judged by the hiring manager against the JD. The trial mechanic and the offer detail are in the operating-partner one-pager linked from the LinkedIn cross-post. Pick the portco. Send the JD. The work starts Monday.