01 · OBJECTIVE
A baseline for the executive cohort
This report establishes a compensation baseline for data, analytics & AI executives at the VP level and above — a global cohort (82% U.S.-based), with all salaries collected in U.S. dollars. It is a companion to the annual Burtch Works Report on salaries of data scientists and AI professionals, extending the same segmentation discipline to the executive cohort those professionals report into. Data was collected through the 2026 Data, AI & Analytics Executive Benchmark Survey — conducted jointly by Burtch Works and CDO Magazine, and fielded May 20 – August 19, 2026. Beyond compensation, the survey benchmarks executive scope, reporting structures, organizational influence, and AI ownership and accountability. Because 2026 is the first edition, no year-over-year comparisons are presented; future editions will trend against this baseline.
Responses are confidential and reported only in aggregate: no company-specific or individual compensation information is published, and the data is used exclusively for this research.
02 · THE CLEANING WATERFALL
From 401 responses to 243 validated executives
Responses remaining at each step
03 · CLEANING & REPORTING RULES
Every exclusion, documented
Cohort definition
Respondents self-classified by degree of organizational oversight. VP and above qualify: VP, SVP/EVP, and C-suite. Directors, managers, and individual contributors were excluded. The cohort is global (82% U.S.-based) and the survey collected all figures in annual USD.
Unit normalization & validity floor
Salary responses entered in thousands (e.g., “350” for $350,000) were scaled by 1,000. Zero, blank, and any base salary below $10,000 were treated as non-response — 5 records, all evident data-entry errors rather than low salaries. This is a data-quality rule, separate from the outlier policy.
Outlier policy — no trimming
Every valid response is retained. 7 base salaries fall outside 1.5× the interquartile range for their level; all are included in the figures reported here and listed individually in the accompanying workbook. Because extreme values are retained, medians are the reliable statistic — means are shown for completeness but are sensitive to the tails.
Total cash & bonus imputation
Unusable total-cash answers (below base or above 3× base) are not dropped — each respondent is retained with an imputed bonus so base salary and total cash share the same sample of 243. 45 of the 47 imputed records use the median bonus rate observed within the target-bonus band the respondent selected; 2 had no band and use their level median. The 9 who answered “No bonus target” are set to zero bonus rather than imputed.
Small-cell suppression
Any segment with fewer than five respondents is shown with counts only. Medians are emphasized over means throughout, and sample sizes are labeled on every table.
Why our research is different
Burtch Works surveys the professionals themselves — not HR departments, not scraped job postings — and this research powers the benchmarks in every retained search we run. Sample sizes and suppression rules are published so you can judge every number in context.
REFERENCES
Sources referenced in this report
Where this report makes a claim about the market that its own survey does not measure, the claim is attributed. Nothing anecdotal appears without a source.
EY Chief Data Officer Role Study
SECOND EDITION, APRIL 2026
Global study of data, analytics and AI leaders across Europe, LATAM, the US and APAC. Cited here for the split between unified and separated data and AI leadership; the finding that 33% of CDOs hold full authority over AI while 80% share it; the shift in titles toward AI-inclusive naming; reporting-line distribution (CIO 23%, CEO 20%, CTO 11%, CFO 11%); and the correlation between organizational structure, role clarity and data-office performance.
Burtch Works search practice
RETAINED EXECUTIVE SEARCHES, 2025–2026
Observations attributed to our own search activity are labelled as practice observations wherever they appear, and are never mixed into survey figures. Every number in the compensation, segmentation, role and organizational-design sections comes from the survey described above. The takeaways section is guidance: it draws on those numbers but adds judgement, and says so.
02 · THE CLEANING WATERFALL
From 401 responses to 243 validated executives
Responses remaining at each step
03 · CLEANING & REPORTING RULES
Every exclusion, documented
Cohort definition
Respondents self-classified by degree of organizational oversight. VP and above qualify: VP, SVP/EVP, and C-suite. Directors, managers, and individual contributors were excluded. The cohort is global (82% U.S.-based) and the survey collected all figures in annual USD.
Unit normalization & validity floor
Salary responses entered in thousands (e.g., “350” for $350,000) were scaled by 1,000. Zero, blank, and any base salary below $10,000 were treated as non-response — 5 records, all evident data-entry errors rather than low salaries. This is a data-quality rule, separate from the outlier policy.
Outlier policy — no trimming
Every valid response is retained. 7 base salaries fall outside 1.5× the interquartile range for their level; all are included in the figures reported here and listed individually in the accompanying workbook. Because extreme values are retained, medians are the reliable statistic — means are shown for completeness but are sensitive to the tails.
Total cash & bonus imputation
Unusable total-cash answers (below base or above 3× base) are not dropped — each respondent is retained with an imputed bonus so base salary and total cash share the same sample of 243. 45 of the 47 imputed records use the median bonus rate observed within the target-bonus band the respondent selected; 2 had no band and use their level median. The 9 who answered “No bonus target” are set to zero bonus rather than imputed.
Small-cell suppression
Any segment with fewer than five respondents is shown with counts only. Medians are emphasized over means throughout, and sample sizes are labeled on every table.
Why our research is different
Burtch Works surveys the professionals themselves — not HR departments, not scraped job postings — and this research powers the benchmarks in every retained search we run. Sample sizes and suppression rules are published so you can judge every number in context.
REFERENCES
Sources referenced in this report
Where this report makes a claim about the market that its own survey does not measure, the claim is attributed. Nothing anecdotal appears without a source.
EY Chief Data Officer Role Study
SECOND EDITION, APRIL 2026
Global study of data, analytics and AI leaders across Europe, LATAM, the US and APAC. Cited here for the split between unified and separated data and AI leadership; the finding that 33% of CDOs hold full authority over AI while 80% share it; the shift in titles toward AI-inclusive naming; reporting-line distribution (CIO 23%, CEO 20%, CTO 11%, CFO 11%); and the correlation between organizational structure, role clarity and data-office performance.
Burtch Works search practice
RETAINED EXECUTIVE SEARCHES, 2025–2026
Observations attributed to our own search activity are labelled as practice observations wherever they appear, and are never mixed into survey figures. Every number in the compensation, segmentation, role and organizational-design sections comes from the survey described above. The takeaways section is guidance: it draws on those numbers but adds judgement, and says so.
REVIEW LOG
What we tested, and what held
This edition was reviewed by an external panel before publication. Each analytical claim raised in that review was reproduced from the record-level data. Claims that did not survive are listed alongside those that did, because a benchmark that only publishes its confirmations is not a benchmark.
CONFIRMED
Median base and total cash cross-cut by revenue band and level
Reproduces exactly from the record-level data, cell for cell.
CONFIRMED
Moving from a 21–30% to a 51%+ bonus target is worth about $306,000 in total cash
We compute $305,964 in median total cash between those two bands.
NOT SUPPORTED
Financial services is the lowest-paying major private industry once scale is held constant
Financial services sits below other private industries only in the under-$1B band. In the two bands where most of its respondents sit it is level or ahead. The raw premium is a scale effect, but the sector does not underpay for an equivalent seat.
PARTIAL
Women in comparable seats are paid 13–18% more than men
Direction holds, magnitude does not. Removing the public sector from both sides moves the female premium from +3.3% to +4.7%, not into double digits. The +13.8% figure is real at VP level only. The composition point behind the claim is confirmed: women here are more than three times as likely as men to hold a public-sector seat.
NOT SUPPORTED
Public companies pay about 12% less base for an equivalent seat
Within revenue band, public and late-stage private medians sit within a few percent of each other in all three bands. The real difference is the equity multiple — 0.93× base at late-stage private against 0.58× at public companies.
PARTIAL
Company revenue is the largest single predictor of base salary
Span of control edges it out, and both trail the bonus target band, which is a component of pay rather than a driver of it. No factor explains more than a fifth of the variation on its own.
ADOPTED
Equity is better expressed as a multiple of base than as a dollar grant
The multiple travels between pay levels in a way the dollar value does not. Every startup-stage cell falls below n=5 and is suppressed.
CONFIRMED
AI scope expansion tracks confidence that the role survives
57% of those whose responsibilities expanded significantly are very confident the role persists, against 33% of those reporting no material change.
SAMPLE SIZES
Respondent counts for every segment
The tables throughout this report show compensation figures without a count column, so they read cleanly. Every count sits here instead. Any segment below three responses is not reported at all; segments between three and four are published but rest on a thin base, and we would treat them as directional rather than as a benchmark.
Title
Industry
U.S. Census region
Country
Company revenue
Organization headcount
Company stage
Reporting line
Work arrangement
Gender
Education
Years of experience
Counts shown in gold sit below five responses. Total cohort: 243 validated executives.
01 · FOR EMPLOYERS
Designing an offer that lands
01
Benchmark on revenue band first, level second
Below $1B the chief title carries almost no premium in this data — VP, SVP/EVP and C-suite medians land at $250,000, $250,000 and $255,000. Anchoring an offer to the title rather than the scale of the organization is how ranges drift.
02
Design the bonus target before the base
47% of the market targets 31%+ of base. Moving a target from 21–30% to 51%+ is worth about $306,000 in median total cash. A 1–10% target performs worse on both pay and satisfaction than no target at all.
03
If you are private, compete on the equity multiple
Base is within a few percent between public and late-stage private companies at the same revenue. Late-stage private grants run 0.93× base against 0.58× at public companies. That is where private wins — say it explicitly in the offer.
04
Budget the negotiation into the range
76% of these executives negotiated, and most who did got more. Approving a range that already contains the negotiation margin costs less than losing a finalist at the last step.
02 · FOR EXECUTIVES
Reading your own market
01
Optimize the bonus target, then the base
Moving from a 21–30% target to 51% or more is worth roughly $306,000 in median total cash. No other single change in this study moves the number that much. Ask what the target is before you argue about base.
02
Below $1B, do not trade cash for a title
The market does not reprice a chief title at that scale. Above $1B it does, and the step is substantial. Know which side of that line the company sits on before treating the title as compensation.
03
Negotiate. A competing offer is not the prerequisite
Median base is identical at $300,000 whether an executive held no competing offer, one, or several. Those who negotiated and got more report $310,000. Waiting for a competing offer is waiting for something that does not move the median.
04
Compare offers at total cash, not base
Median total cash is $400K against $300K in base, and the spread widens with seniority. Two offers with the same base can be $200,000 apart.
One caution
The group that negotiated and got nothing reports a median base $50,000 below the group that simply accepted the first number. We read that as selection rather than penalty — executives in weaker positions negotiate harder and still lose — but it is a real pattern and worth knowing before assuming negotiation is free.
03 · THE MARKET BY SEGMENT
Where the mandate differs most
Ownership structure changes both the pay and the job. Public companies (42% of the cohort, $312.5K median) buy scale and governance; late-stage private including PE-backed (27%, $300K) buys value creation on a clock; government and nonprofit (15%, $270K) carries the clearest discount. The private-equity themes below are where we see the sharpest version of the new executive mandate — but they are one segment of several, and the commentary reflects Burtch Works search activity alongside the survey data.
01
The playbook is evolving toward operational value creation
Financial engineering and multiple expansion still matter, but returns increasingly come from how portcos perform day-to-day: gross profit, EBITDA margin, durable enterprise value. Many firms now field fund-level operating teams — sometimes larger than their investment teams.
02
AI is rapidly becoming a core lever for value creation
No longer experimental — a foundational driver of productivity, efficiency, and margin expansion. PE leans in because it is incentivized by hold periods, structurally faster than public companies, and actively seeking partners who translate AI strategy into execution.
03
Data is a strategic advantage at the fund level
Leading firms invest in interoperable data infrastructure across portcos, centralized analytics, and decision intelligence informing both deal flow and operations. Aggregating and operationalizing portfolio-wide data compounds performance and speed.
04
The talent model is going forward-deployed
Demand is rising for data & AI leadership at fund and portco level, operators who implement rather than advise (repeatable ROI use cases), and flexible talent models — contract and project-based included. The premium sits on business acumen, execution, and ownership.
05
Riches in niches — specialization compounds
Firms are doubling down on sector specialization: packaging industry-specific AI use cases that redeploy across portcos and feed the underwriting process pre-acquisition as part of the value-creation thesis.
06
Where the momentum is
Healthcare (data scale and complexity, transformation upside), lower-middle-market companies (operational upside, earlier in AI adoption), and VC-to-PE transitions, where firms operationalize and scale previously venture-backed businesses.
Hiring an AI operating partner or forward-deployed talent for a portfolio? See our private equity search practice.
04 · THE SEAT ITSELF IS CHANGING
AI is expanding the executive mandate
91%
Say AI expanded their responsibilities in the past 12 months
86%
Are confident their role exists in its current form in 3 years
51%
Are satisfied with total compensation — 25% report dissatisfaction
How AI changed role scope
Confidence the role persists (3-year view)
Retention context: one in four executives reports some dissatisfaction with total compensation, 41% of hires arrived with competing offers, and a fifth of the cohort is less than a year into the seat. In a market still creating seats faster than it fills them, the leaders you already have are being benchmarked — whether you do it or someone else does.