This is one of the hardest organisational questions in the industry right now, and most agencies are getting it wrong in one of two predictable ways: either they treat technologists as a service function attached to "the real work," or they spin up a separate tech team that's structurally walled off from strategy and creative. Neither produces the integrated thinking the work actually requires. The agencies getting this right are doing something more uncomfortable, rebuilding the org around the assumption that strategy and technical capability are now the same job at the senior level, even when they're different people.
The starting point is recognising why the old separation existed and why it doesn't anymore. For decades, agencies were organised around three crafts that genuinely required different skills, different temperaments, and different career paths: account management (relationships and orchestration), strategy (insight and positioning), and creative (ideas and craft). Production was a service layer underneath, and "tech", when it existed, was either a small digital studio doing banner ads or an outsourced relationship with a development shop. This worked because the deliverables were artifacts: a campaign, a TV spot, a print ad, a website. Artifacts have clear handoffs. You could brief them across functions and stitch the result together at the end.
What broke this was the shift from artifacts to systems. A modern marketing engagement isn't a campaign you ship, it's a system you operate: an always-on personalisation engine, a creative variant pipeline, a measurement infrastructure, a data activation layer, an AI workflow that ingests briefs and produces outputs continuously. Systems can't be briefed across silos and stitched at the end, the technical decisions made in week one constrain what's strategically possible in month six, and the strategic intent has to be encoded into the system's design from the start. When strategy and engineering are separated, you get systems that work technically but miss the point, or strategies that are conceptually sound but can't actually be built. Both failure modes are common, and both come from the same root cause: the people who understood the goal and the people who understood the mechanism weren't in the same conversation early enough.
So the real question isn't "how do we add technical people to the agency?", it's "how do we restructure so that strategic and technical thinking happen together from the first meeting." A few patterns are emerging from the agencies that have been trying this for the last few years.
The first pattern is the hybrid senior role, where a single person carries both the strategic and technical authority for an account or a domain. These titles vary, "principal strategist," "technology strategist," "solution architect," "growth engineer", but the function is the same: someone senior enough to be in the room when business strategy is being set, who also has the technical depth to know what's actually buildable, what it will cost, what the second-order effects are, and what the team needs to start doing differently. These roles are extremely hard to hire for because they're a genuine intersection skill, not just two skills stacked. The people who can do them well usually came up through one discipline and deliberately learned the other over years, a strategist who taught themselves data engineering, or an engineer who got pulled into client work and discovered they were good at it. They're rare, expensive, and usually the linchpin of whatever's working in an AI-native agency. The mistake most agencies make is trying to manufacture them by pairing a strategist and an engineer and calling it a hybrid role; that's a partnership, which is useful, but it isn't the same thing as one mind holding both perspectives.
The second pattern is the integrated pod, where the unit of delivery is a small cross-functional team that includes strategy, creative, technology, and account leadership from day one of an engagement. This is the most common structural answer, and the one that usually looks right on a slide but is hard to make work in practice. The friction points are predictable: the technical team's velocity and the creative team's process have different rhythms, the strategy team wants to think before doing while the engineering team wants to prototype before deciding, the account team needs predictable deliverables to manage the client relationship while the technical team needs space to discover what the right answer actually is. These tensions don't go away by putting everyone in the same Slack channel, they need explicit operating agreements about how decisions get made, how disagreements get resolved, who has authority over what, and what the team's actual unit of output is. Pods that don't do this work end up replicating the silos at a smaller scale, with everyone in the same room but still passing briefs across an invisible wall.
The third pattern is the embedded specialist, where technical people aren't pooled in a central team but are embedded directly into strategy or creative groups. The advantage is that technical thinking gets injected into every conversation rather than being something you "consult the tech team about." The disadvantage is that embedded specialists can lose touch with their craft community, there's nobody to learn from, nobody to push their thinking, nobody to share the latest tools and approaches with. The agencies doing this well pair embedding with a strong cross-cutting practice: the data engineer is embedded in the brand strategy team for client work, but participates in a weekly engineering practice forum, has a senior technical mentor, and is evaluated on craft growth as well as account contribution. Without that craft anchor, embedded specialists tend to either drift into junior strategist roles (losing their technical edge) or burn out and leave.
The fourth pattern, less common but worth naming, is the parallel practice with explicit interlocks, where strategy and technology run as separate practices but with formalised rituals for working together, joint kickoffs, shared OKRs, paired leadership, mandatory technology review at strategic milestones. This works best at larger agencies where the scale justifies separate practice areas, and it's often more honest than the integrated pod model because it doesn't pretend the cultural differences aren't real. The risk is that the interlocks become bureaucratic theatre, everyone shows up to the joint meeting, nobody actually changes how they work, and the silos persist underneath the rituals. The agencies that make this work treat the interlocks as load-bearing, not ceremonial.
Across all of these patterns, a few principles seem to separate the agencies that are blending talent well from the ones that are just claiming to.
Hire for translation, not just craft. The bottleneck isn't usually finding strategists or finding engineers, it's finding people on either side who can translate between the two. A strategist who can read a data model and understand its implications is worth two who can't. An engineer who can sit in a brand strategy session and understand what business outcome the work is supposed to drive is worth two who think the brief is "just give me the requirements." Agencies that hire for this translation capability, explicitly screening for it in interviews, paying for it, promoting on it, build teams that work across disciplines naturally. Agencies that hire pure specialists on both sides spend a lot of energy on integration overhead.
Pay technical talent on the actual market, not the agency salary band. This sounds obvious and is constantly violated. A good data engineer or ML practitioner has Big Tech, finance, and martech vendors competing for them, and the agency's traditional salary structure, designed around account directors and creative directors with relatively narrow market comparables, doesn't compete. Agencies that try to fit technical talent into the existing band lose the best people within eighteen months and end up with the ones who couldn't get hired elsewhere. Agencies that bite the bullet and pay technical talent at market rates, even when it creates internal pay disparities with strategists at the same level, build durable technical capability. The pay disparity is uncomfortable and politically expensive, but the alternative is worse.
Give technical leaders real authority, not just consultation rights. A common failure mode is hiring senior technical talent and then routing all their decisions through a non-technical account or strategy lead who can override them. This is demoralising and produces bad outcomes, the technical person either learns to give the answer the senior person wants to hear (which defeats the point) or leaves. The agencies getting this right give technical leaders explicit authority over technical decisions: what platforms get used, what gets built vs. bought, what the data architecture looks like, how engineering time is allocated. The strategic leadership sets direction; the technical leadership owns execution and has the authority to push back when the direction isn't buildable as scoped.
Train the strategists, not just the engineers. Most agency upskilling efforts focus on teaching creatives and strategists "how to use AI tools", which is necessary but not sufficient. The deeper investment is teaching strategy and account people enough about how the technology actually works that they can make competent decisions when there isn't an engineer in the room: what a CDP actually does, what a model context window is and why it matters, what the difference is between deterministic and probabilistic matching, what an API rate limit means for a campaign timeline, why "let's just have AI do it" is sometimes wrong. This isn't about turning strategists into engineers, it's about closing the gap enough that the strategists can be useful collaborators rather than people who hand off requirements and hope for the best. The agencies that invest in this seriously change the texture of every cross-functional conversation. The ones that don't keep having the same misunderstandings indefinitely.
Create a culture where technical people can shape the strategy, not just implement it. This is the hardest cultural shift, because it requires senior strategists and creatives to accept that their ideas will get genuinely challenged by people from a different discipline — not in the "well actually, technically we can't do that" gotcha way, but in the substantive "the technical reality of how this would actually work changes what the right strategic answer is" way. The agencies that allow this become smarter; the agencies that defend the strategy hierarchy from technical input keep doing strategy that doesn't survive contact with implementation. A test for whether this is working: in your last big strategic decision, did a technologist meaningfully change the conclusion? If the answer is no, they're not actually in the room, they're just sitting in the chair.
A few honest observations about how hard this is in practice.
The cultural distance between traditional agency talent and engineering talent is large, larger than people coming from outside the industry usually appreciate. Agency culture is built around presentation, persuasion, judgment, and craft expressed through taste. Engineering culture is built around iteration, evidence, reproducibility, and craft expressed through systems that work. Both are real crafts and both produce excellent work, but they're not the same craft, and the people who do them well don't always immediately respect each other. Bridging this isn't a structural problem, you can't org-chart your way out of it. It's a leadership problem about modelling mutual respect and creating shared experiences of solving real problems together. The agencies that lead this from the top do okay; the ones that delegate it to HR don't.
The timeline is longer than leadership wants to hear. Building a genuinely integrated practice, not just hiring some technologists, but rewiring how the agency thinks and decides, typically takes 3-5 years from a serious start. Most agencies don't have the patience for this, partly because the holding companies they belong to don't have the patience for it, and partly because the leadership team that committed to the change in year one isn't always still there in year four. The agencies that have actually pulled this off usually had a CEO or managing partner who personally championed it for the long haul, made the unglamorous investments, weathered the early years when it looked like it wasn't working, and didn't blink when the first hires didn't pan out. There's no shortcut.
Some of the work needs to leave. This is the part nobody wants to say in pitch meetings, but it's true. Trying to be excellent at strategy, creative, technology, and engineering all at once at every scale doesn't work for most agencies. The ones that succeed usually make hard choices about what they're going to be world-class at and what they're going to either partner for or genuinely walk away from. An agency that tries to keep its old creative production capacity intact while also building a serious data engineering practice is going to underfund both. The agencies thriving in 2026 mostly look more focused than they did in 2018, fewer disciplines, more depth, and explicit partnerships for the things they no longer try to do in-house.
The deeper point is that the strategic-and-technical question isn't really about hiring or org structure, it's about what an agency is fundamentally for. The old agency was a vendor of artifacts: you brief them, they make the thing, they ship it. The new agency, when it works, is a partner in building and operating systems that drive the client's business. That requires a different kind of team, but more importantly it requires a different kind of relationship with the client, a different commercial model, and a different self-conception. Agencies that have made the technical hires without making the deeper shift end up with engineers sitting around waiting for someone to brief them on the next deliverable. Agencies that have made the deeper shift find that the engineers, the strategists, and the creatives are all working on the same problem from the start, and that the work is meaningfully different, and meaningfully better, than what any of them could have produced separately.
