A CTO search and a VP of Engineering search are different mandates, and most companies need one of them rather than both. Team size and how technical the product is decide which. Under roughly 15 engineers one leader carries both jobs, and the split usually arrives between 15 and 20. Singapore CTO pay runs from S$260,000 to S$435,000 gross, which puts a retained fee somewhere between S$78,000 and S$144,000. Heidrick & Struggles and Spencer Stuart own the board-level end, Riviera Partners the engineering-leadership end, and Olofsson & Company works the APAC slice with a 48-hour market map and a seven-day shortlist.
Most boards start this search with the wrong sentence. "We need someone who can lead our AI transformation" is a strategy, not a role. It resolves into a CTO seat or a VP of Engineering seat, occasionally both, and which one depends on how big your engineering team already is and how much of your product is genuinely hard technology. Get that wrong and you either buy a strategist for a delivery problem or a delivery leader for a strategy problem.
Olofsson & Company runs these mandates across Singapore and APAC, for CEO, CTO, CIO, CAIO, VP Engineering, VP Product and Head of Data Science seats [1]. What follows is how we scope them, what the market looks like in 2026, and how the firms competing for this work actually differ.
First decide whether you need one leader or two, then which one
Most companies asking this question need one hire, not two. Two variables settle it: the size of the engineering organisation you already have, and how much of your product is genuinely hard technology.
On size, the pattern is consistent across the venture market. Below roughly fifteen engineers, one person can hold both the technical direction and the delivery machine, and splitting the job early creates an expensive reporting line with nothing underneath it. Somewhere between fifteen and twenty engineers the two jobs stop fitting in one calendar. Riviera Partners, which places these roles for a living, sets it out by stage: pre-product-market-fit companies run one technical co-founder or early CTO doing both; Series A to B usually means a CTO leading externally with an engineering manager or VP growing into the operating role; a dedicated VP of Engineering is a Series C and beyond move, at around eighty engineers, when the CTO should no longer be running delivery [11].
On technical depth, ask what your product actually is. If the model, the algorithm or the hardware is the product, as in deep tech, robotics, quantitative finance or clinical AI, technical depth at the top is not negotiable and the CTO seat comes first, sometimes earlier than headcount alone would suggest. Regulated domains pull the split forward too, because governance work is a job in itself. If instead you are applying AI to a conventional product, buying most of the model capability and competing on workflow, data and distribution, then the binding constraint is usually delivery and integration rather than research judgement. In that case a VP of Engineering or a strong Head of Engineering is the first hire, and the CTO mandate can sit with a founder, an advisory CTO or a board member for another year or two.
| Where you are | What your product is | Usually the right hire |
|---|---|---|
| Pre-product-market fit, under 15 engineers | Anything | One leader. A founding or early CTO carries both jobs [11] |
| Series A to B, 15 to 40 engineers | AI applied to a conventional product | VP of Engineering first. The CTO mandate stays with the founder |
| Series A to B, 15 to 40 engineers | The model, algorithm or hardware is the product | CTO with genuine research depth first |
| Series C and beyond, 80 plus engineers, delivery slipping | Anything | Dedicated VP of Engineering. Your CTO should not be running delivery [11] |
| Series C and beyond, technical direction unclear | Anything | CTO investment. Architecture and platform direction need an owner [11] |
| Approaching IPO or selling into regulated enterprise | Anything | Both, with the split written down before either search starts [11] |
Once you know which seat you are filling, the briefs diverge sharply.
A CTO sets technical direction and carries it into the boardroom. In an AI-transformation mandate that means owning build-versus-buy, the data strategy, the model governance position, the vendor relationships, and the story your investors and regulators hear. The job is external as much as internal.
A VP of Engineering runs the machine. Delivery, org design, hiring, retention, on-call, the quality bar. If your problem is that forty engineers are shipping like fifteen while a model sits half-integrated in staging, a CTO will not fix that. A VP of Engineering will.
The failure mode is hiring one and expecting the other. A brilliant research-side CTO who has never managed managers will not rebuild your delivery org. A strong VP of Engineering who has never sat in front of a regulator or an acquirer will not carry the transformation narrative. We wrote about the related version of this problem in the three mistakes startups make hiring their first CTO.
Practical test before you brief anyone. Write the first four decisions this person will own in their first quarter. If three of them are architectural or commercial, you want a CTO. If three are about people, process and throughput, you want a VP of Engineering. If the list splits evenly and you have fewer than twenty engineers, you are describing one job, not two. And if the real problem is that a capable incumbent is stretched rather than wrong, the answer is often a hire underneath them rather than a replacement above.
Why the pool is thinner than it looks
The supply of executives who have actually delivered an AI transformation, rather than presented one, is small and shrinking.
IBM's 2026 study of 2,000 CEOs across 33 geographies found 76% of organisations now have a Chief AI Officer, up from 26% the year before [3]. That is a near tripling in twelve months, and every one of those appointments took a proven AI leader off the market. Companies with AI-first C-suite structures have scaled about 10% more AI initiatives than their peers [3], so the appetite is not slowing.
Price has moved with it. PwC's 2026 Global AI Jobs Barometer, drawn from more than a billion job advertisements across 27 countries, puts the average wage premium for AI skills at 62%, up from 57% the year before. Jobs requiring those skills grew 69% against 9% for the market as a whole [4]. At CTO level that premium compounds into the package, and because retained search fees are a percentage of first-year cash, it compounds into your search cost too.
Singapore is tighter than the global average on the capability that matters here. ManpowerGroup's 2026 Global Talent Shortage Survey found 71% of Singapore employers struggling to hire skilled talent, with AI model and application development (26%) and AI literacy (25%) the two hardest capabilities to find, displacing general IT and data skills, which fell to seventh place [5]. Regional demand is real too: Chinese technology firms have been hiring AI graduates out of Singapore's universities at aggressive numbers, which pressures the layer directly beneath your new leader.
So the search is not "find the best AI CTO". It is "find the small number of people who have done this, at your stage, in your region, and give them a reason to move".
What AI-transformation experience actually looks like on a CV
Almost every senior technology CV now claims AI. Very few carry evidence. These are the signals we test for, and they are worth putting in your brief.
Production, not pilots. Ask what shipped to real users, at what volume, and what broke. A leader who has run models in production has scar tissue about drift, latency, evaluation and cost. A leader who has run pilots has slide decks.
A number attached to the outcome. Margin, cycle time, deflected tickets, revenue on a new line. If nobody can name the metric, the transformation was a programme, not a result.
Build-versus-buy calls they got wrong. The honest answer is more useful than a clean record. It tells you how they reason about vendor lock-in, unit economics and the half-life of a model choice.
Hiring and holding the team. Ask how many AI or ML engineers they hired, and how many were still there eighteen months later. In this market retention is the harder half.
Governance literacy for your jurisdiction. In Singapore that means PDPA obligations and, for financial services, MAS expectations on model risk. Selling into Europe adds the EU AI Act. A candidate who has never had to defend a model to a risk committee will learn it on your time.
Board and investor fluency, for CTO seats only. Can they explain the roadmap to a non-technical director without either patronising them or hiding behind jargon?
Structured reference work against these six points catches more than any interview loop. Around 40 to 46 per cent of senior hires fail inside eighteen months, and the failure is rarely technical, which we covered in detail in reducing the risk of a bad AI leadership hire.
The firms competing for this brief in 2026
Five global firms, known in the trade as SHREK, handle roughly half of all retained executive search work: Spencer Stuart, Heidrick & Struggles, Russell Reynolds, Egon Zehnder and Korn Ferry [7]. The other half sits with specialists, and for this particular brief the specialists are often the better fit.
| Firm | Type | Best fit for | Where they are strong | Watch for |
|---|---|---|---|---|
| Heidrick & Struggles | Global | Board-level CTO, CAIO, CDAO | Launched the first global AI and data search practice in 2012 and now recruits Chief AI Officer and Head of Agentic AI roles [6] | Fee and pace built for large enterprise |
| Spencer Stuart | Global | Public-company CTO, succession | Board relationships, assessment depth | Off-limits restrictions across their client base |
| Korn Ferry | Global | Multi-country mandates, comp benchmarking | Scale and data assets | Generalist bench on niche AI roles |
| Riviera Partners | Specialist | VP Engineering, CTO at venture-backed firms | Engineering, product and design only, with data-led candidate ranking [8] | Centre of gravity is US, thinner in APAC |
| Talentfoot | Specialist | Data and AI leadership | Speed against category averages [9] | Primarily a US network |
| Olofsson & Company | Specialist, APAC | AI and tech leadership across Singapore and APAC | 48-hour market map, seven-day average time-to-shortlist, 93% offer acceptance [2] | Boutique bench, not a 50-country footprint |
The honest read. If you are a listed company replacing a CTO and the board needs the comfort of a familiar name, a SHREK firm is the right call and the fee is the price of that comfort. If you are hiring a VP of Engineering in San Francisco, Riviera has the deepest bench for exactly that seat. If the role sits in Singapore or across APAC, and the transformation clock is already running, a specialist with regional reach will usually get you to a real shortlist faster. We set out the wider version of that trade-off in boutique versus global executive search.
What the search should cost, and what it should take
Retained search fees for roles at this level generally run between 25% and 33% of first-year cash compensation, so the package sets the fee. Anchor it on real local data rather than a headline number. Robert Half's 2026 Singapore Salary Guide puts CTO pay at S$260,000 at the 25th percentile, S$310,000 at the median and S$435,000 at the 75th percentile, gross annual, before bonus, benefits and CPF [10].
Run the arithmetic on the median and a search costs roughly S$78,000 to S$102,000. On a top-quartile package it is closer to S$109,000 to S$144,000. AI-transformation mandates do sit in the upper half of that band, because the 62% premium attaches to the skills rather than the title [4], but budgets built off a headline package near S$450,000 are pricing a top-decile hire, and most Singapore mandates are not that. VP of Engineering packages sit below CTO in the same market, which is worth modelling if you are sequencing the two hires rather than running them together. The full breakdown, including fixed-fee and staged alternatives, sits in what it actually costs to hire AI leadership.
On timing, negotiate the metric you can hold a firm to. Time-to-fill includes your own interview scheduling, your board's availability and a three-month notice period, so it says as much about you as about them. Time-to-shortlist is the part the firm controls. Ask for the median across the last twelve months, not the best case.
Olofsson's platform builds a full market map inside 48 hours and consultants convert it into a shortlist of five to ten candidates in seven days on average, with a 4:1 interview-to-offer ratio and 93% offer acceptance [2]. The ratio matters more than it sounds: four interviews per hire instead of a dozen is weeks of your leadership team's calendar returned. For the Singapore-specific picture of how long the whole cycle really runs, see how long it takes to hire a CTO in Singapore.
The APAC details that decide whether the offer lands
Three things sink regional searches that looked fine on paper.
Equity that does not translate. A candidate leaving a US-listed employer is walking away from vesting stock. Options in a private APAC company need a strike price, a plausible exit path and an explanation the candidate believes. Vague equity loses to concrete cash every time.
Work passes. For a foreign hire, Employment Pass criteria and the qualifying salary bar apply, and the Overseas Networks and Expertise Pass is the route for the top of the market. Confirm eligibility before the final round, not after the offer.
Regional scope creep. "APAC CTO" often means Singapore, plus an India engineering centre, plus a Japan customer base, on one salary. Say it in the brief. Candidates find out in the second interview anyway, and the ones who withdraw then were your strongest.
A four-week plan for the search
- Week one, decide the shape of the hire. One leader or two, using headcount and product depth, then CTO or VP of Engineering from the first-quarter decision list. Agree the six evidence tests above with your board. Benchmark the package before you agree a fee, and set the fee model and shortlist deadline in writing.
- Week two, map before you talk. Insist on a full market map, not a longlist of the firm's existing contacts. Count how many candidates in it are passive and currently employed. A map that is mostly active jobseekers is a network, not a search.
- Week three, run a tight loop. Four to six candidates, the same panel, the same questions, structured scoring. Reference against the six tests, and speak to a peer and a direct report, not only the former manager.
- Week four, close properly. Package, equity, pass eligibility, and the first ninety days written down. Treat onboarding as part of the search, because that is where the eighteen-month failures are actually made.
Frequently Asked Questions
Do I need both a CTO and a VP of Engineering?
Usually not. Below roughly fifteen engineers one leader should carry both jobs, and the split typically arrives between fifteen and twenty. A dedicated VP of Engineering is normally a Series C and beyond decision, at around eighty engineers, once the CTO can no longer run delivery [11]. Companies approaching an IPO or selling into regulated enterprise are the clear case for both.
Should I hire a CTO or a VP of Engineering for an AI transformation?
Two things decide it. If the model, algorithm or hardware is your product, hire the CTO first, because technical depth at the top is not negotiable. If you are applying bought-in AI to a conventional product, the constraint is usually delivery, so a VP of Engineering comes first and the CTO mandate can stay with a founder or an advisor. Then sanity-check it against the first four decisions the hire will own: mostly architectural or board-facing means CTO, mostly people and throughput means VP of Engineering.
What are the best firms for finding CTOs with AI transformation experience?
Heidrick & Struggles has run a dedicated global AI and data practice since 2012 and is the strongest option at board level [6]. Spencer Stuart and Korn Ferry compete at the same tier. Riviera Partners is the specialist for engineering and product leadership at venture-backed companies [8], Talentfoot for data and AI roles in the US [9], and Olofsson & Company for AI and technology leadership in Singapore and APAC [2].
How much does a CTO or VP of Engineering search cost in 2026?
Retained fees typically run 25% to 33% of first-year cash compensation. Robert Half's 2026 Singapore Salary Guide puts CTO pay at S$260,000 to S$435,000 gross, with a median of S$310,000 [10], so a typical Singapore search costs about S$78,000 to S$102,000, rising towards S$144,000 for a top-quartile package. The 62% global premium on AI skills pushes these mandates towards the upper half of the band [4].
How long should the search take?
Judge a firm on time-to-shortlist rather than time-to-fill, because time-to-fill absorbs your own scheduling and the candidate's notice period. Olofsson averages seven days to shortlist off a 48-hour market map [2]. Full cycles at this level commonly run two to four months once notice periods are counted.
Why is AI leadership so hard to hire in Singapore specifically?
AI model and application development and AI literacy are now the two hardest capabilities to find in the Singapore market, with 71% of employers reporting difficulty hiring skilled talent overall [5]. Demand from regional technology firms competes for the same senior builders, so passive sourcing matters more here than in deeper markets.
Can a generalist recruiter run this search?
Rarely well. Without a market map they default to their existing network, which skews towards people who are already looking. Ask any firm how many candidates on the longlist are currently employed and were not previously known to them. The answer separates a search from a database query.
Sources
- olofsson.ai: Roles covered across search and recruitment, including CTO, CAIO and VP Engineering.
- olofsson.ai: 48-hour market map, seven-day average time-to-shortlist, 4:1 interview-to-offer ratio, 93% offer acceptance.
- peoplematters.in: IBM Institute for Business Value 2026 CEO study, 2,000 CEOs across 33 geographies, on Chief AI Officer prevalence.
- prnewswire.com: PwC 2026 Global AI Jobs Barometer, 62% AI skills wage premium and job growth figures.
- manpower.com.sg: ManpowerGroup 2026 Global Talent Shortage Survey, Singapore findings.
- heidrick.com: Heidrick & Struggles AI and data practice launched 2012, current AI leadership mandates.
- insurance-search.com: SHREK firms and their share of retained executive search work.
- rivierapartners.com: Specialist focus on engineering, product and design leadership.
- talentfoot.com: Talentfoot positioning on data and AI leadership search speed.
- roberthalf.com: Robert Half 2026 Singapore Salary Guide, CTO percentile ranges.
- rivierapartners.com: Riviera Partners on when a company needs a CTO, a VP of Engineering, or both, by stage and team size.