Last updated: June 30, 2026

How to Hire a VP of Data: 2026 Guide

By Robert Ardell, Co-Founder and Strategic Advisor, KORE1

Hiring a VP of Data in 2026 means budgeting roughly $200,000 to $300,000 in base salary plus equity, and deciding first whether you are consolidating several scattered data teams under one executive or just renaming a manager. Most of these searches run eight to fourteen weeks. The ones that go sideways were lost in the first meeting, before anyone wrote a job description, when nobody settled what the role was actually for.

I have been part of KORE1 since we started placing data and technology leaders in 2005, and I will tell you where I sit before I tell you anything else. We run executive searches for data leaders and we staff the data engineering and data science teams that report up to them, and we only get paid when you hire. So when this guide tells you to slow down, or that what you are describing is a director and not a VP, that advice sometimes argues against my own invoice. Read it that way. A wrong VP of Data hire costs more than any fee I would send you, and it keeps costing you for two years.

The title is the problem. “VP of Data” gets written on a req by a forty-person startup and a four-thousand-person bank in the same week, and they mean completely different people. One needs someone who will still open a notebook. The other has not written SQL since 2019 and should not start now. Three words. Two different people.

VP of Data, Head of Data, or Chief Data Officer?

Answer this before anything else. It decides your budget, your candidate pool, and whether the person you hire is bored or drowning by month three. Get it wrong and you pay twice.

Here is the distinction that gets blurred. A head of data builds and runs one data team, often the first one a company has ever had, and frequently still writes code. A VP of Data runs the whole data function through other managers, usually after the company already has a few data teams that grew up in different corners and now need a single owner. A chief data officer sits at the executive table and owns what data is worth to the business at the board level. Governance, risk, monetization, and strategy. If you want that person, we wrote a separate chief data officer hiring guide, and most companies reading this do not need them yet.

The clean test for a VP of Data specifically: are you consolidating? If your analytics team reports into finance, your data engineers sit under platform, and a couple of data scientists hang off the product org, and you have finally decided that this should be one function with one budget and one person accountable, that is a VP of Data hire. If you are hiring your very first data leader into an empty room, you almost certainly want the head of data, and you should go read that guide instead. I would rather lose you to the right page than sell you the wrong search.

One caution on the table. These tiers slide around by company. A “head of data” at a 3,000-person enterprise can outrank a “VP of Data” at a Series B startup, because the enterprise team is forty people and the startup team is nine. Titles are relative. Scope is what you are actually buying. So describe the scope in the req. Stop trusting the noun.

What a VP of Data Actually Owns

A VP of Data owns the output of the entire data function, the people who produce it, and the platform it runs on. That is the one-sentence version. Underneath it sit three jobs that do not look like each other.

The first job is running a multi-discipline org. Data engineering, analytics and BI, data science, and increasingly the data platform and ML infrastructure, all under one person who has to make them feel like one team instead of four tribes that resent each other. This is the part founders underestimate. Badly. A VP of Data spends more time getting a Looker analyst and a Snowflake engineer to agree on what “active user” means than doing anything you would recognize as technical. Politics, mostly. Dressed as architecture.

The second job is architecture and org design, and this is where the VP earns the title. They make the calls a head of data running one team never has to: centralize the data org or embed analysts inside each business unit. Build on Snowflake or Databricks. Standardize on dbt and Airflow or buy a managed stack. Fix data quality before promising the CEO a single AI feature, or ship the feature and accept the technical debt. These are six-figure decisions that compound for years, and they are the actual reason you are paying VP money instead of senior-manager money. Get one wrong and you live with it for three.

The third job is the one nobody screens for and everybody needs. Translation. Turning “we want to be data-driven” from a board into a roadmap a team can build, and turning what the team finds back into something an executive will act on before the quarter ends. A VP of Data who cannot do this becomes a very expensive reporting service. You will know fast. Two quarters, maybe less.

Watch which of the three a candidate gets excited about. The org-design obsessive can build you a beautiful platform that answers questions nobody asked. The translator can charm a board and ship dashboards but quietly let the pipelines rot. The strongest hires are uncomfortable in at least one of the three and honest about which one. The candidate who claims to be equally great at all of it has usually not done the job long enough to know better.

What a VP of Data Costs in 2026

Pull two salary sites and you will think one of them is lying. They are not. They are measuring different jobs that share a title.

ZipRecruiter puts the average VP of Data and Analytics base near $178,900 in early 2026, with most landing between $133,000 and $213,000 and top earners around $261,500. Glassdoor reports an average closer to $275,000, with its upper quartile north of $370,000. That is a gap of nearly $96,000 in the average alone, for the same title. The spread is the whole lesson. ZipRecruiter leans toward base pay at smaller companies. Glassdoor folds in the equity and bonus that funded and public companies pile on top. You are never hiring the average. Nobody is.

For context on the broader market, the U.S. Bureau of Labor Statistics does not track “VP of Data” as an occupation, which tells you how young the title still is. The nearest proxy it does track, computer and information systems managers, earned a median of $171,200 in May 2024 and is projected to grow 15% through 2034. The people a VP of Data manages are scarcer still: data scientist roles are projected to grow 34% over the same decade, the fourth-fastest of any occupation the BLS measures. You are hiring a leader for a team that is getting harder to staff every year. That scarcity is baked into the number.

Here is the band I actually open a budget conversation with, by company stage. Base salary for U.S. hires in major markets, with total comp folding in bonus and equity. Outside the top metros, take 10 to 15% off.

Two notes before you anchor on any of this. First, these assume a permanent, direct hire. If you are genuinely unsure whether the role should exist yet, a fractional or contract-to-hire data leader costs more per hour and far less in regret, and for a first attempt at this seat that trade is often the smart one. Second, a single city’s range is enormous, so a national average will lie to you in both directions. If you want a live read for your exact market and stack, our salary benchmark assistant will pull a current band in about a minute.

The Hiring Process, Step by Step

Six steps. The early ones are boring and everyone rushes them. The late ones are where the rushing shows up, usually around week ten, usually as a great candidate you cannot close because nobody decided what you were offering.

1. Scope the org before you write the job description

Write down which teams report to this seat on day one and which you expect to add by day three hundred. Data engineering only? Engineering plus analytics? The whole thing including data science and ML platform? This single answer changes the salary, the seniority, and the kind of person who will say yes. I watched a healthcare company in the Midwest run a nine-week VP search, fall for a brilliant platform architect, and realize too late that the job was 70% stakeholder politics across three business units the architect had no patience for. He lasted seven months. Scope first. Everything downstream inherits this decision.

2. Set the comp band to 2026 reality

Decide base, bonus, and equity before you talk to a single candidate. People at this level negotiate compensation for a living, and if you walk in without a number they will set it for you, high. Pull the band for your stage and your metro, not the national average that blends a Series A startup against Capital One. The CDO salary guide is worth a glance here too, because at the enterprise level a VP of Data and a CDO start to overlap on pay, and you should know where your offer sits relative to the title above it.

3. Decide how you are going to run the search

Three honest options. Your own network, an in-house recruiter, or a retained search firm. Networks are free and fast when they work and silent when they do not. In-house recruiters are excellent at filling engineering roles and out of their depth on a VP search they run maybe once a year. Retained search exists for exactly that gap, which is why our retained executive search practice handles most of our data-leadership hires. A VP of Data is a permanent direct-hire placement, not a contractor, so the whole search is built around someone who will still be in the chair in three years. I will tell you below when to skip all of this and hire on your own.

4. Source where data executives actually are

They are not on job boards. Never the strong ones. The good ones are employed, busy, and getting recruited quietly every month, which means the search is outbound or it is nothing. Where they actually surface: running data orgs one size below yours, speaking at Snowflake Summit or the local Databricks and dbt meetups, and writing the thoughtful kind of post that no keyword search will ever catch. The best signal is not a resume. It is a leader whose former reports keep following them to the next company. That loyalty is the job itself, already proven. Chase that.

5. Run a loop that tests org-building, not SQL

Throw out the take-home coding test. You are not hiring a senior engineer, and a VP who needs to prove they can write a window function is the wrong VP. Test the three things the job is made of instead. Give them a real, messy situation from your actual org, like “analytics and data engineering disagree on who owns the metrics layer,” and watch them resolve it out loud. Have them sketch a 90-day plan for a team your size and listen for what they refuse to promise. Then put them in a room with one of your business stakeholders, because a VP of Data who cannot win over the head of finance will be ignored by exactly the people whose data they are supposed to fix.

6. Make the offer, then protect the first 90 days

Once you are sure, close quickly. Strong data leaders are usually weighing two or three other conversations at the same time, and the company that drags its feet is the one that loses them. So put the real terms in writing on day one. The mandate, the budget, the reporting line, and the first three problems you actually want solved. A senior candidate reads a vague offer as a vague job, and a vague job is the easy one to walk away from. Then plan the first quarter together. Give them one early win the executives will notice, the dashboard they keep asking for or the broken number everyone argues about, and let them bank that credibility before you bury them in the backlog.

How to Tell a Real Data Executive From a Good Résumé

At this level every résumé is a highlight reel. They all “built a data-driven culture,” all “scaled the team,” all “delivered AI.” So the paper tells you almost nothing. The real work is separating the person who drove the outcome from the one who happened to be in the room when it landed.

A few signals I trust more than the bullet points. Look for someone who has consolidated teams that did not want to be consolidated, because merging analytics-under-finance with engineering-under-platform is a knife fight, and surviving one teaches things no greenfield build ever will. Ask about a tool or platform they killed. Real data leadership is mostly deciding what to stop doing. Mostly. A candidate who has never sunset a beloved internal tool or migrated off a stack everyone defended has not made a hard call yet.

Ask who they hired and where those people are now. A real data executive leaves a trail of analysts and engineers who went on to run their own teams. That trail is the actual deliverable, and it is the hardest thing to fake on a résumé. We ran a search a few years back where the eventual hire had the least flashy project list of the three finalists. What she had instead was a former intern now running analytics at a company twice the size of her last one, and two data engineers who had followed her across two job changes because they refused to work for anyone else. The client wanted the finalist with the splashy AI launches on his résumé. We pushed for her. Three years on, half their data org traces back to people she hired and grew, and the splashy candidate had already left his next job. People who build people are rare. When you find one, the résumé will almost always undersell them.

And mind the operator who has only ever worked at one size of company. A VP who ran a 60-person data org at a public company can be helpless at a 25-person startup, where there is no platform team to delegate to and the VP is the platform team for a while. The reverse breaks too. Stage fit is not a tiebreaker here. It is most of the decision.

When You Can Run This Search Yourself

I run a search firm, so weigh this accordingly. You do not always need me. I mean that. I would rather say so than sell you a search you can run on your own.

If your head of data has been quietly doing the VP job for a year, the teams already follow them, and the only thing missing is the title and a wider mandate, promote them. Skip the fee, keep the cash, and spend it on the headcount they have been asking for. Internal promotions into this seat work more often than founders expect, because the hardest part of the job is context, and an internal candidate already has years of it. Context is the moat. You cannot hire it in from outside. The mistake I see is the opposite one. A founder assumes the candidate with the famous logo on the résumé must be better than the person already doing the work, pays a premium plus a six-month ramp, and learns the hard way that pattern recognition from a company ten times your size does not always translate down.

Where a retained search earns its fee is the hire you cannot afford to miss and cannot reach on your own. A first VP of Data consolidating three feuding teams. A confidential replacement where the current leader does not yet know they are being replaced. A turnaround after the last VP failed and the team has gone gun-shy. Those land on our desk, because the cost of getting them wrong dwarfs any fee, and because pulling a senior data leader out of a job they are not even looking to leave is a specialized kind of work. We have done it since 2005, our placements hold at a 92% twelve-month retention rate, and the recruiters on these searches average more than fifteen years apiece. If the stakes are high and you cannot run the search quietly on your own, that is the one to hand off.

What Founders and CTOs Ask Us Before Hiring a VP of Data

VP of Data or head of data, what’s the real difference?

A head of data builds and runs one team, usually your first. A VP of Data runs several teams through managers, usually after the company already has data people scattered across the org.

The cleanest tell is consolidation. If you are bringing analytics, engineering, and data science under one roof for the first time, that is a VP. If you are hiring your very first data leader into an empty room, that is a head of data, and you will get a better result and a lower bill reading the head of data guide instead. The titles slide by company size, so always describe the scope, not just the noun.

Should the data org be centralized under the VP, or embedded in the business units?

Centralize first, then selectively embed. A brand-new VP of Data needs central control to fix data quality and set standards before handing analysts back out to the business.

The fully embedded model, sometimes dressed up as a data mesh, works at companies with mature data platforms and strong governance already in place. Most companies hiring their first VP of Data have neither yet. Embedding analysts before the foundation exists just scatters the mess across more teams. Centralize to build the standards, then push analytics back toward the business units once the plumbing is trustworthy. This is exactly the kind of call you are hiring the VP to make, so ask them how they would sequence it.

Do we need a VP of Data if we already have a CDO?

Often yes. A CDO sets strategy and faces the board; a VP of Data runs the teams that execute it. At larger companies they are two different jobs, not one.

The CDO owns governance, risk, and what data is worth to the company. The VP owns delivery: pipelines, reporting, models, and the people who build them. In a company under about a thousand people, one strong leader can wear both hats. Past that, splitting them is usually what unblocks the function, because the board-facing work and the team-running work pull in opposite directions and starve each other when one person tries to do both.

What does a VP of Data cost us, all in?

Plan on $190,000 to $300,000 in base salary for 2026, with total compensation reaching $420,000 to $620,000 at growth and pre-IPO companies once bonus and equity stack.

Base is the predictable part. Equity is where two VPs with identical titles end up six figures apart, and it is your strongest lever when cash is tight at an earlier stage. Aggregator averages swing by nearly $100,000 for this title because they blend startups and public companies, so benchmark against your own stage and metro, not a national number you will have to defend to finance later.

Where should a VP of Data report?

Most report to the CTO, the CDO, or the CEO. The reporting line matters less than the budget and hiring authority that come with it.

Up to the CTO is the common default, and it works when the CTO actually values data instead of treating it as a feature request. Up to a CDO makes sense once the strategy and execution layers are split. Up to the CEO signals that data is a first-class priority and helps you land stronger candidates. What sinks the role is parking it three levels deep under a CIO. Strong candidates spot that setup in the first interview and quietly take themselves out of the running.

How do we know our head of data is ready to be a VP?

When they are already managing managers, making platform and budget calls, and spending most of their week on people and strategy instead of pipelines, the title is just catching up to the job.

The risk is promoting your strongest individual contributor into a role that is 70% leadership, because the skills do not automatically carry. Promote the head of data who already mentors, already translates for the executive team, and actually wants to manage at a larger scale. Not the one who is simply your most technical person. If you are unsure, that read is exactly what an outside data staffing partner can give you before you commit.

Get the Scope Right and the Rest Follows

Every hard part of this hire traces back to the first question. Are you consolidating an org, or staffing your first team? Answer it honestly, write the req for the answer, pay the real 2026 number for your stage, and give whoever you hire the budget and authority to do the job you described. Do that and this is a clean search. Skip it and no amount of sourcing saves you.

If you are about to pull your data teams under one leader and you want an outside read on whether this is really a VP seat or something a step below, that is a conversation we have most weeks. Talk to our executive search team. We will give you the honest version, including the one where you promote from within and never send us a dollar.