Analytic Partners review

Strong — 4 of 5 on this rubric’s scale, from Poor (1) to Excellent (5). This dimension carries 25% of the total score.

Benchmark — across the 16 other agencies evaluated in this discipline, this dimension runs Excellent 1 · Strong 3 · Adequate 12. The typical agency here scores Adequate, and 1 of them score higher than this one.

What this dimension measures: The heaviest weight, because an unvalidated model is an opinion with decimal places. Look for a described methodology, stated assumptions, and above all how the model is VALIDATED: holdout periods, backtesting, or reconciliation against a real experiment. A described method with no validation story scores 3.

Scores high — A firm that publishes its validation approach and names its limits scores 4-5.

Scores low — 'Proprietary algorithm' with no methodology at all scores 1-2 — proprietary is not a method, and in this category it is the single least checkable claim a vendor can make.

What we found — A published blog names the actual validation vocabulary rather than hiding behind proprietary language: backtesting, holdout testing described as 'reserving a time period from training to test against afterward - also called out-of-sample testing', confidence intervals, and an argument that these are 'necessary but not sufficient on their own' with real-world prediction validation as the test 'most vendors skip'. The MMM page names a concrete limit of the standard method - models that omit macro conditions, competitive investment, distribution and pricing 'systematically assign excess credit to media' - and cites 45 percent of an ad's impact as cross-channel halo. A whitepaper, 'Calibrating with Chaos', is specifically about contamination in experiments used to calibrate models. Held below Excellent because the model form is never stated, no validation results for their own models are published, the calibration whitepaper is gated behind a download form, and the solution pages for Commercial Analytics, MMM and Pricing Optimization carry no validation language at all. ROI Genome is described as proprietary intelligence. That is the high band above, which is why it scored Strong.

On the record — “Validation vocabulary is published rather than hidden behind proprietary language: backtesting and holdout/out-of-sample testing are named and described, and stated to be 'necessary but not sufficient on their own', with real-world prediction validation named as the test most vendors skip.” analyticpartners.com ↗

On the record — “The Q1 2026 Forrester Wave Leader placement is reported by independent trade press as well as by the firm, but the report itself is gated behind the vendor's own download form.” itbrief.asia ↗

Evidence: vendor stated — the agency’s own claim, recorded as theirs rather than ours.

Strong — 4 of 5 on this rubric’s scale, from Poor (1) to Excellent (5). This dimension carries 20% of the total score.

Benchmark — across the 16 other agencies evaluated in this discipline, this dimension runs Excellent 1 · Strong 6 · Adequate 5 · Weak 4. The typical agency here scores Adequate, and 1 of them score higher than this one.

What this dimension measures: Whether the firm can establish causality rather than only correlation: geo holdout tests, matched-market design, PSA/ghost-ad tests, switchback designs, or reconciliation of modelled results against live experiments.

Scores high — Named experiment designs with described execution score 4-5.

Scores low — Correlation-only modelling with no experimental capability scores 2-3 — legitimate and common, but the buyer should know they are buying a correlational estimate. A firm that presents modelled attribution as proven causality scores 1-2, and the overclaim should be named in the verdict.

What we found — A dedicated Incrementality Testing solution page names designs rather than gesturing at them: 'geo-, store-, and audience-based test designs - from matched-market and holdout tests to multi-cell experiments'. Execution is described (tests designed and run inside GPS-E, duration forecast at the design stage, 'timelines vary by test design and scale, typically running several weeks to allow for reliable signal'), as is the controlling-for step - adjusting for seasonality, competitor activity, weather and category shifts to reduce bias and contamination. Test results are stated to feed back into the forecasting and planning models, which is the reconciliation loop this dimension asks for, and the gated calibration whitepaper addresses the risk in doing exactly that. PSA/ghost-ad and switchback designs are not named. Not Excellent: no executed test is published with a named client, a design and a result. That is the high band above, which is why it scored Strong.

On the record — “Named experiment designs are published: 'geo-, store-, and audience-based test designs - from matched-market and holdout tests to multi-cell experiments', with typical run times of several weeks and results fed back into the GPS-E forecasting and planning models.” analyticpartners.com ↗

attribution — working out which marketing touch actually caused a sale. Good practice names its model and its blind spots; bad practice quotes each ad platform’s self-graded numbers, which overlap and overclaim.

Evidence: vendor stated — the agency’s own claim, recorded as theirs rather than ours.

Strong — 4 of 5 on this rubric’s scale, from Poor (1) to Excellent (5). This dimension carries 15% of the total score.

Benchmark — across the 16 other agencies evaluated in this discipline, this dimension runs Excellent 1 · Strong 8 · Adequate 6 · Weak 1. The typical agency here scores Strong, and 1 of them score higher than this one.

What this dimension measures: Attributable client work at a stated scale and category. Treat any 'we found X% waste' claim as a vendor-stated number unless a client is named and corroborates it.

Scores high — Named clients with described engagements score 4-5.

Scores low — Anonymised case studies score 2-3 — common here for genuine confidentiality reasons, so do not penalise beyond the band, but do not credit unverifiable lift figures either.

What we found — Six named customer stories are published (Cox Communications, Hill's Pet Nutrition, Nespresso, Colgate-Palmolive, VF Corporation, plus an ADAPTA data-onboarding story), and the homepage names Beiersdorf, Cox, Kroger, VF Corporation and Nespresso; the MMM page carries a SharkNinja quote. The Cox story is a real described engagement - 2021-2022, scenario planning, channel-level CPM monitoring, a performance waterfall dashboard - not a logo wall. Held at Strong because every outcome number is the vendor's own: 10 percent marketing efficiency improvement at Cox, 30M dollars of potential revenue in the ADAPTA story, 25-70 percent ROI gains on the platform page, 25 billion dollars of opportunity supported. No client is quoted by name and title in the Cox study, and none of the lift figures is corroborated by the client. That is the high band above, which is why it scored Strong.

On the record — “Six named enterprise customer stories are published, including Cox Communications, Hill's Pet Nutrition, Nespresso, Colgate-Palmolive and VF Corporation.” analyticpartners.com ↗

On the record — “The Cox Communications engagement is described with a concrete cadence - a dashboard updated every four weeks with quarterly monitoring - and a vendor-stated 10 percent marketing efficiency improvement between 2020 and 2022, with no named client quote.” analyticpartners.com ↗

CPM — cost per thousand ad impressions — the unit price of reach.

Evidence: vendor stated — the agency’s own claim, recorded as theirs rather than ours.

Strong — 4 of 5 on this rubric’s scale, from Poor (1) to Excellent (5). This dimension carries 15% of the total score.

Benchmark — across the 16 other agencies evaluated in this discipline, this dimension runs Strong 10 · Adequate 4 · Weak 2. The typical agency here scores Strong, and none scores higher than this one.

What this dimension measures: What the engagement needs from the buyer (spend, conversion, and channel data at what granularity) and — critically — whether the firm also buys the media it is measuring. A firm that measures media it does not sell is structurally more credible; where the same firm buys and grades its own work, that conflict must be disclosed and should be named in the verdict whether or not the firm names it.

Scores high — Clear data requirements plus independence from media buying scores 4-5.

Scores low — Undisclosed conflict scores 1-2.

What we found — Independence from media buying is the strongest fact here and it is corroborated off their own site: their private equity owner Onex describes the business as a platform that helps customers 'optimize future allocations across offline and online media channels' - it measures media it does not sell, so the buy-and-grade-your-own-work conflict does not apply. On data, the Commercial Analytics page makes an unusually specific and checkable commitment - 'Our Commercial Analytics solution doesn't require user-level data. It uses a multi-dimensional structure that incorporates customer segments or cohorts' - and states that fragmented sources (retail media networks, co-op programs, partnership channels) are integrated rather than modelled in isolation. Held below Excellent because no required input list or granularity is published: a buyer cannot tell what spend, conversion and channel data they must supply, at what level, or over what history. One residual tension to note: the About page lists measurement partnerships with Amazon, Google, Meta, Microsoft, TikTok, Pinterest and Roku - the platforms whose media the firm grades. That is the high band above, which is why it scored Strong.

On the record — “The engagement does not require user-level data: 'Our Commercial Analytics solution doesn't require user-level data. It uses a multi-dimensional structure that incorporates customer segments or cohorts.'” analyticpartners.com ↗

On the record — “The firm does not sell the media it measures. Its owner Onex describes it as a cloud-based managed software and service platform that helps F1000 customers assess marketing spend effectiveness and optimize future allocations across offline and online media channels; Onex has held the investment since 2022.” onex.com ↗

Evidence: partly checkable — corroborated in part against the sources below; the remainder rests on the agency’s own account.

Adequate — 3 of 5 on this rubric’s scale, from Poor (1) to Excellent (5). This dimension carries 10% of the total score.

Benchmark — across the 16 other agencies evaluated in this discipline, this dimension runs Excellent 1 · Strong 6 · Adequate 9. The typical agency here scores Adequate, and 7 of them score higher than this one.

What this dimension measures: What arrives and how often: a one-off model, a refreshed quarterly model, a live dashboard, a decision workshop. A model delivered once and never refreshed is a snapshot of a market that has moved, and should be scored as the ceiling it is.

Scores high — Stated deliverables with a stated refresh cadence score 4-5.

Scores low — Undefined deliverables score 2.

What we found — The platform page states what arrives - scenario simulations, forecasts, optimization recommendations and multi-dimensional insight in a single dashboard, built from ADAPTA (ingestion and cleansing), AMP (the analytics engine) and PROPHET (forecasting, planning, war-gaming, optimization) - and the model is continuous rather than a one-off study ('always-on models and insights', 'real-time course corrections'). But no standard refresh cadence is published on any solution or platform page. The only concrete cadence located anywhere on the site sits inside the Cox case study: a dashboard updated every four weeks with quarterly monitoring. A buyer reading the offer pages cannot tell how often their model is rebuilt or what the standing deliverable calendar is. That is between the two bands, which is why it scored Adequate.

On the record — “The Cox Communications engagement is described with a concrete cadence - a dashboard updated every four weeks with quarterly monitoring - and a vendor-stated 10 percent marketing efficiency improvement between 2020 and 2022, with no named client quote.” analyticpartners.com ↗

On the record — “No pricing, fee range, engagement structure, minimum, or post-engagement model licence terms appear on the homepage, platform page, any of the six solution pages read, or the Contact Us page.” analyticpartners.com ↗

deliverability — deliverability: whether cold email actually lands in the inbox rather than spam. The tell is infrastructure talk — warmed sending domains kept separate from your main domain — because a burned domain outlasts the engagement.

Evidence: vendor stated — the agency’s own claim, recorded as theirs rather than ours.

Weak — 2 of 5 on this rubric’s scale, from Poor (1) to Excellent (5). This dimension carries 10% of the total score.

Benchmark — across the 16 other agencies evaluated in this discipline, this dimension runs Strong 2 · Adequate 3 · Weak 11. The typical agency here scores Weak, and 5 of them score higher than this one.

What this dimension measures: Note any minimum spend and whether the model licence continues after the engagement ends.

Scores high — Published fees, ranges, or a stated engagement structure with numbers score 4-5.

Scores low — A described structure without numbers scores 2-3. Bespoke-only with no anchor scores 1-2.

What we found — Measured absence across homepage, /platform/, /solutions/, the MMM, Incrementality and Commercial Analytics pages, and /contact-us/ - no fee, no range, no engagement structure, no minimum, no statement of whether the model licence survives the engagement. The Contact Us page offers only a discussion and a demo. The one cost anchor found anywhere is indirect and inside a study the firm itself commissioned: the Forrester TEI composite carries three-year risk-adjusted service fees of 2.0M dollars present value for an organization with a 350M dollar annual marketing budget. Third-party dollar ranges exist on competitor comparison blogs; those are interested parties and are not credited here. That is what the low band describes, which is why it scored Weak.

On the record — “No pricing, fee range, engagement structure, minimum, or post-engagement model licence terms appear on the homepage, platform page, any of the six solution pages read, or the Contact Us page.” analyticpartners.com ↗

Evidence: inferred — our reading of indirect evidence, not a documented fact.

Stronger here: Nepa scores Strong on the same dimension.

Adequate — 3 of 5 on this rubric’s scale, from Poor (1) to Excellent (5). This dimension carries 5% of the total score.

Benchmark — across the 16 other agencies evaluated in this discipline, this dimension runs Adequate 5 · Weak 11. The typical agency here scores Weak, and none scores higher than this one.

What this dimension measures: Independent, verified reviews or checkable references. Low weight deliberately: measurement work sells through procurement and referral, so a thin public review footprint is normal and the method evidence above matters far more.

Scores low — A substantial verified base scores 4-5; a handful scores 2-3; none located scores 2.

What we found — The independent evidence that exists is analyst-firm evaluation rather than client reviews: named a Leader in The Forrester Wave: Marketing Measurement and Optimization Services, Q1 2026, ranked highest in Current Offering with top scores on 21 of 31 criteria - reported by trade press (IT Brief Asia) as well as by the firm, though the report itself is gated behind the vendor's form. A 2025 Gartner Magic Quadrant Leader placement is claimed but was not read first-hand. Both Gartner Peer Insights and the G2 product page returned HTTP 403 to first-hand retrieval, so no rating or review count is cited here. A thin public client-review footprint is normal for procurement-led measurement work, which is why this dimension carries the lowest weight.

On the record — “The Q1 2026 Forrester Wave Leader placement is reported by independent trade press as well as by the firm, but the report itself is gated behind the vendor's own download form.” itbrief.asia ↗

On the record — “The site is not a soft-404 site: /this-page-cannot-possibly-exist-9f3k2 returned HTTP 404, so page content read on this domain matches the URL requested.” analyticpartners.com ↗

Evidence: partly checkable — corroborated in part against the sources below; the remainder rests on the agency’s own account.