Elissa Brown, ecommerce evangelist and Industry Lead at AppsFlyer, used her appearance at Shoptalk Spring 2026 to argue that organizational structu...
The Signal
Elissa Brown, ecommerce evangelist and Industry Lead at AppsFlyer, used her appearance at Shoptalk Spring 2026 to argue that organizational structure—not technology—is the primary reason measurement programs fail. Speaking to ClickZ at the event, Brown identified cross-team silos as the single largest gap between what measurement platforms deliver and what marketing organizations actually use. Social, investment, competitive intelligence, and planning teams, she noted, routinely generate insights for the same brand in parallel rather than in coordination, fragmenting the measurement picture regardless of the sophistication of the underlying tools. Brown's position carries operational weight beyond the advisory: she is simultaneously building Farmer's Community Branding, a CPG food company, from the ground up, giving her direct exposure to measurement failures at both startup and enterprise scale. On attribution, she stopped short of declaring last-click dead, instead framing it as a signal that crowds out others when used as the default methodology—pushing teams toward bottom-funnel optimization at the expense of upper-funnel demand creation. Her recommended corrective is incrementality testing alongside last-click, rather than wholesale replacement.
Why It Matters
The real problem Brown is diagnosing isn't measurement maturity—it's organizational design, and most martech vendors have a financial incentive to keep that conversation off the table. Selling a better attribution tool is a tractable sales motion. Telling a CMO their team structure is broken requires a different conversation entirely, which is why it rarely happens. Brown can say it because AppsFlyer's credibility is built on measurement outcomes, not software licenses alone—and because she has the founder receipts to back it up. What surface-level coverage of this interview misses is the competitive implication of the silo problem. When social, investment, and planning teams generate parallel insights, the brand doesn't just lose measurement coherence—it loses the ability to learn faster than its competitors. In a landscape where incrementality testing and multi-touch models are increasingly accessible to mid-market brands, the differentiator isn't access to the methodology. It's the organizational wiring to act on it in a coordinated way. The brands running integrated measurement cycles are compounding their learning advantage every quarter. Brown's incrementality-alongside-last-click framing is also more strategically sophisticated than it sounds. She's not just offering a methodological middle ground—she's describing a change management approach. Ripping out last-click creates internal resistance and data continuity problems. Layering incrementality next to it builds the organizational muscle for full-funnel thinking without triggering the defensive reflex that kills most measurement reform efforts inside large teams. For senior decision-makers, the uncomfortable question this surfaces is whether your current measurement investment is solving a data problem or covering for a people and process problem. If it's the latter, another platform evaluation won't fix it.
The Exploit
Opportunity
Mid-market brands with a CMO willing to treat org design as a competitive lever—not just a people problem—can open a measurable learning gap on slower competitors within two quarters. The window is now: incrementality testing infrastructure is newly accessible at non-enterprise price points, meaning the differentiator through Q3 2026 is execution speed, not tooling access.
Risk
Brands that respond to measurement underperformance with another platform evaluation are compounding the wrong problem. The failure mode is recognizable: a new attribution vendor gets selected, reporting improves cosmetically, and bottom-funnel bias persists because the social, investment, and planning teams are still running parallel insight cycles with no shared accountability structure. Competitors who solve the org problem while you solve the software problem widen their learning advantage every sprint.
The Move
Before Q3 2026 planning locks headcount and budget, the CMO should commission a single cross-functional audit—owned by the head of marketing operations, not a vendor—mapping where social, investment, and planning teams currently generate measurement outputs and whether any shared review cadence exists. If no cadence exists, that is the capital allocation decision: fund the coordination layer before the next tool.
First-Mover Advantage
Gains
Risks
Brands that respond to measurement underperformance with another platform evaluation are compounding the wrong problem. The failure mode is recognizable: a new attribution vendor gets selected, reporting improves cosmetically, and bottom-funnel bias persists because the social, investment, and planning teams are still running parallel insight cycles with no shared accountability structure. Competitors who solve the org problem while you solve the software problem widen their learning advantage every sprint.
Window
Winners & Losers
Winners↑
Incrementality Testing Vendors
Brown's explicit endorsement of incrementality-alongside-last-click as a change management entry point—rather than a wholesale replacement—creates a durable commercial opening for vendors who can position their tools as additive rather than disruptive to existing workflows. The mechanism is risk reduction: brands that resist ripping out last-click can adopt incrementality testing without triggering internal resistance, lowering the sales barrier significantly. Vendors in this space should lead Q2 and Q3 2026 messaging with the 'layer on, not replace' framing to convert fence-sitters at mid-market and enterprise accounts.
Organisational Design and Marketing Operations Consultancies
Brown's core argument—that measurement fails because of team structure, not technology—creates direct demand for the consultancy category that martech vendors structurally avoid selling. The mechanism is a conversation gap: because platform vendors have a financial incentive to frame every problem as a tooling problem, the organisational redesign conversation is systematically undersupplied, leaving a clear lane for consultancies that can diagnose and fix cross-team silo problems. Firms in this space should use Brown's Shoptalk framing as third-party validation to reposition measurement engagements as org design engagements starting now.
Full-Funnel Media Planning Agencies
The silo critique Brown articulates—social, investment, competitive intelligence, and planning teams generating parallel insights for the same brand—describes the exact coordination failure that integrated agency models are built to solve. Agencies that unify channel planning, measurement, and reporting under one roof gain a structural advantage as brands recognise that fragmented team outputs are costing them learning velocity, not just data coherence. The response opportunity is to make compounding learning speed a headline capability claim in new business pitches through Q3 2026, with incrementality cadence as the proof point.
The Marketing Exploit
By surfacing the organisational design layer that surface-level coverage of Brown's interview missed, The Marketing Exploit positions itself as the outlet that reads competitive implications rather than just reporting statements—directly reinforcing its differentiation from commodity event recap journalism. The mechanism is editorial authority: when the actual strategic insight is 'measurement investment may be covering for a people and process problem,' only outlets willing to say that uncomfortable thing build trust with senior decision-makers. Continuing to anchor original analysis to under-covered structural implications of widely-covered industry events is the compounding editorial advantage to press in Q2 and Q3 2026.
Losers↓
Single-Methodology Attribution Platforms
Platforms whose core value proposition is built on last-click attribution as a default—rather than as one signal within a broader measurement mix—face an accelerating credibility problem as the incrementality-alongside framing gains mainstream endorsement from credible practitioners like Brown. The mechanism is positioning erosion: when a respected AppsFlyer industry lead says last-click 'crowds out other signals' and drives bottom-funnel over-optimisation, it shifts the burden of proof onto vendors who have not evolved their methodology narrative. Defensive action requires reframing their offering around signal orchestration rather than single-source attribution before this framing hardens into conventional wisdom through 2026.
Siloed In-House Performance Marketing Teams
Brown's diagnosis lands as a direct indictment of the organisational structures many in-house teams have calcified around—where social, paid search, and planning functions report separately, measure separately, and optimise separately without a shared measurement cycle. The mechanism of pressure is competitive compounding: as Brown notes, brands running integrated measurement cycles build a learning advantage every quarter, meaning siloed teams are not just inefficient today but structurally slower learners over time. The defensive play is to use Brown's Shoptalk profile to build internal business cases for measurement integration—framing the org redesign conversation as a competitiveness argument rather than an operational one.
Strategic Outlook
The timing here isn't accidental. Shoptalk Spring has increasingly become the venue where operational reality checks get aired—less product announcement, more practitioner reckoning. Brown's argument lands at a moment when the measurement vendor landscape has consolidated significantly, with AppsFlyer, Rockerbox, and Northbeam all competing for mid-market incrementality budgets that barely existed three years ago. The tools got democratized faster than the organizations using them matured. That gap is now visible in outcomes. The silo problem Brown is naming has a longer lineage than most coverage acknowledges. The same structural critique surfaced during the last-touch attribution wars of 2019 and 2020, when MTA vendors promised unified views and quietly discovered that the data integration problem was really a politics-of-attribution problem in disguise. Most of those vendors are gone or absorbed. The methodology moved on; the org chart didn't. What's different in 2026 is the stakes. With retail media networks fragmenting measurement further—Amazon, Walmart Connect, and Instacart each running attribution on their own rails—the cost of siloed teams compounds quarterly. Brands that haven't wired cross-functional measurement loops by Q3 2026 will enter peak season with learning deficits their competitors have already closed. That's the window Brown is really pointing at.