Boathouse’s New CSO Says Agencies Must Build Their Own Measurement Rails—Or Keep Renting Them From Big Tech
While Madison Avenue waits for Google to hand out post-cookie measurement gifts, one midsize shop just hired its own referee. In March, Boston-based independent Boathouse quietly named Sonia Chung its first-ever Chief Strategy Officer—a move that signals indie agencies are re-orging around data ownership, not media leverage.
Chung’s mandate is simple: reclaim the yardstick before the rulers disappear. With Privacy Sandbox APIs still shifting underfoot and Amazon’s clean-room sprawl tightening its data grip, whoever owns the pipes owns the performance narrative—and the fee floor.
“Agencies need to focus on expanding their capabilities and developing their own infrastructures,” Chung told AdExchanger in her first interview since taking the role. Translation: stop outsourcing measurement to the same platforms that sell you inventory.
From Digitas to Google to Salesforce—Now Back to the Agency Side
Chung’s hybrid résumé is the currency Boathouse is betting on. She began on the agency side at Digitas, steering Fortune 100 accounts through the first wave of programmatic. She then jumped to Google just as Ads Data Hub was being cobbled together, and later landed at Salesforce during the CDP gold rush. That tour of duty gives her the political capital to sell clients a short-term “tax”—higher upfront data-infrastructure spend—in exchange for a long-term “tariff removal” once platform CPMs lose their measurement moat.
“Everybody was optimizing their channels to death, but those channels each represent a piece of a much larger pie,” she says. “What’s missing is connecting and activating on the data.”
AI as the Always-On Planning Intern
Boathouse’s answer to the connective-tissue problem is a three-layer AI stack:
- Creative versioning engine – every resized banner or 6-second cut-down becomes a micro-experiment.
- Media optimizer – spend shifts across search, social, and CTV before quarterly rebate talks even begin.
- Strategy simulator – gen AI acts like “another person in the room,” pressure-testing hypotheses at sprint speed.
“I always talk about gen AI like it’s another person, in addition to our teams, to pressure-test certain ideas and strategies,” Chung explains. The result: a self-built, channel-agnostic outcome graph that lives inside Boathouse’s own cloud tenancy—sandbox-free by design.
The agency is already seeing early dividends. One financial-services client reduced cost per funded account by 18 percent after the AI surfaced that retargeting dollars were cannibalizing high-intent organic search traffic. The model moved $400 k out of paid social and into branded search within two weeks—no platform rep negotiation required.
Turning Waste Into Signal
Traditional agencies treat creative versioning as a tax on the production budget; Chung treats it as a data dividend. Every additional headline or color swap is logged against downstream KPIs—application starts, account opens, even lifetime-value curves—then fed back into the model.
“AI allows some of that to happen faster and can see through some of the complexities to connect those dots,” she notes. In other words, what used to be landfill (unused banners) is now signal (measurable micro-conversions).
The approach also insulates Boathouse from the coming creative-deprecation wave. Once third-party cookies are fully deprecated, look-alike pools inside Google or Meta will shrink, making first-party creative performance the only scalable targeting lever left. Boathouse’s asset-level feedback loop means its training data gets richer exactly when platform data gets poorer.
The Short-Term Pain Pitch
Convincing finance-minded CMOs to fund bespoke data pipes isn’t easy. Chung’s workaround: frame infrastructure as media arbitrage.
- Year 1: client spends an extra 5–7 percent of media budget on cloud warehousing and identity scaffolding.
- Year 2: agency negotiates direct publisher deals and clean-room matches, cutting DSP and SSP take rates by roughly the same margin.
- Year 3: once cookies are gone, platform CPMs lose their measurement moat; Boathouse clients pocket the delta while competitors scramble for Google’s latest “privacy-safe” black box.
It’s the agency equivalent of installing solar panels: higher capex today, lower opex tomorrow—and immunity to the utility’s rate hikes.
Ripple Effects Across the Holding Companies
Boathouse is a 200-person shop, but its experiment is a proxy war for the holding companies. If Chung’s bet pays off, the next Publicis or Omnicom earnings call may be answering to boutiques that stopped renting measurement—and started printing it.
Independent media advisors are already taking note.
“We’re seeing RFPs that explicitly ask, ‘Show us your first-party infrastructure roadmap,'” says one media-consultancy CEO who asked not to be named. “Six months ago that line item didn’t exist.”
Even the platforms are hedging. Google’s latest Ads Data Hub update now lets agencies export aggregated reach curves—a quiet admission that clients want custody of their own telemetry. Amazon, meanwhile, is pitching universal ID clean rooms to agencies willing to co-model retail and advertiser data, but only if the agency brings its own identity spine.
The 90-Day Sprint Ahead
Chung’s first 90 days are already booked:
- Audit every client measurement contract—identify where platform tags can be replaced with server-side events.
- Stand up a unified taxonomy—search keywords, social hashtags, and CTV placement IDs must map to the same conversion object.
- Negotiate data-processing rights—publishers get higher share of wallet in exchange for raw log-level feeds.
If she succeeds, Boathouse will have done more than hire a CSO; it will have issued a proof-of-concept for the post-platform agency. And every other indie shop still waiting for Google’s next measurement gift will be left holding a very expensive rental agreement.
💡 Deep Dive: Don’t miss our Ultimate Industry Guide for advanced strategies.
