Analytic Edge review

Adequate — 3 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 4 · Adequate 11. The typical agency here scores Adequate, and 5 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 — Method is described at workflow level only. The Demand Drivers page names a six-step sequence (Input, Review, Modelling, Reporting, Simulation, Optimization) but publishes no holdout protocol, no backtest description, no stated assumptions and no named limits. The one validation artifact I could read is a blog post arguing MMM should be calibrated against lift experiments, which states that fit was assessed via R-square and MAPE and claims 2 in 3 Meta MMM ROI results changed after calibration, averaging a 25 percent shift; the underlying method sits in a gated white paper I could not read. Working credit for the March 2025 Google certification to implement Meridian, an open-source and publicly documented MMM framework, which is more checkable than a proprietary black box. Capped at Adequate under the corroboration rule: the validation approach itself is not published and the limits are not named, and the flagship Demand Drivers model remains proprietary. That is between the two bands, which is why it scored Adequate.

On the record — “The firm publishes the position that MMM output should be calibrated against lift experiments, claiming that two in three Meta MMM ROI results changed after calibration with an average 25 percent shift, and citing R-square and MAPE as the fit measures used. The underlying method sits in a gated white paper that could not be read.” analytic-edge.com ↗

On the record — “Demand Drivers is described as a six-step workflow (Input, Review, Modelling, Reporting, Simulation, Optimization) offered as full-service, SaaS, or SaaS plus FTE, ingesting data via API with manual UI load for ad hoc inputs. No validation method, holdout, backtest, stated assumption, limitation, defined refresh frequency or price appears on the page.” analytic-edge.com ↗

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

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 — SynTest is a named, separately documented in-market testing product built on synthetic control methodology, with four named designs: matched market / geo testing, lead market geo testing, matched store testing, and matched audience testing, plus stated built-in QA and sensitivity metrics. Notably the firm does not present modelled attribution as proven causality; its published position is the opposite, that MMM output needs calibration against lift experiments. The Codeway engagement records the client moving on to Meta Conversion Lift studies after the model. Execution detail beyond the design names (power calculations, control selection criteria, minimum test duration) is not published, which is why this is vendor-stated rather than verified. That is the high band above, which is why it scored Strong.

On the record — “SynTest is a documented in-market testing product using synthetic control methodology, supporting four named designs: matched market / geo testing, lead market geo testing, matched store testing, and matched audience testing, with stated built-in QA and sensitivity metrics.” analytic-edge.com ↗

On the record — “The firm publishes the position that MMM output should be calibrated against lift experiments, claiming that two in three Meta MMM ROI results changed after calibration with an average 25 percent shift, and citing R-square and MAPE as the fit measures used. The underlying method sits in a gated white paper that could not be read.” analytic-edge.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 — Of 21 case studies on the site, most are anonymised by sector (multinational beverage company, QSR retailer, Canadian bank, Middle East supermarket chain). Two are named with described engagements: Codeway and Melia. The Codeway study is specific and carries an attributed quote from a named individual, Nazli Dagdelen, Performance Marketing Manager, and reports Meta at 31 percent of media-driven new subscribers, rising to 44 percent with halo effects, and a 30 percent fall in cost per subscription. Those figures are vendor-stated and I credit none of them as outcomes. Scale is corroborated indirectly: the March 2025 Google press release states 3000-plus MMMs delivered over three years, and the 2020 Meta engagement covered four Facebook advertising clients. That is the high band above, which is why it scored Strong.

On the record — “Of 21 case studies, two name the client with a described engagement (Codeway, Melia); the rest are anonymised by sector. The Codeway study describes weekly model results plus a secondary halo-effect model, carries an attributed quote from Nazli Dagdelen, Performance Marketing Manager at Codeway, and reports Meta at 31 percent of media-driven new subscribers rising to 44 percent with halo effects and a 30 percent decrease in cost per subscription. Those figures are vendor-stated.” analytic-edge.com ↗

On the record — “Google certified Analytic Edge as an APAC partner for Meridian, Google's open-source MMM platform, announced 4 March 2025, in a consulting and modelling partner role for Google advertisers. The release states the firm has delivered 3000-plus MMMs globally over the past three years and quotes co-founder and director Santosh Nair.” prnewswire.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 15% of the total score.

Benchmark — across the 16 other agencies evaluated in this discipline, this dimension runs Strong 11 · Adequate 3 · Weak 2. The typical agency here scores Strong, and 11 of them score 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 clean: the full solutions list is analytics only (dynamic forecasting, Demand Drivers, PriceSense, Price Pack Architecture, SynTest, revenue growth management, data management, visualization, managed services, loyalty, assortment optimization, integrated planning). Nothing on the site offers media planning, buying, or agency-of-record work, so this firm does not grade media it sold. Data requirements, however, are barely specified: the Demand Drivers page says data arrives by API with manual UI load for ad hoc inputs, but publishes no required fields, no granularity, no history length and no readiness checklist. One relationship worth naming: the firm is a certified Meridian partner for Google and was a 2020 MMM partner to Meta, so it is measuring channels sold by platforms that certify and refer it. That is not a media-buying conflict, but it is not perfect neutrality either, and the site does not disclose it as a consideration. That is between the two bands, which is why it scored Adequate.

On the record — “The full published solutions list contains no media planning, media buying or agency-of-record service, so the firm does not buy the media it measures. Solutions are dynamic forecasting, Demand Drivers MMM, PriceSense, price pack architecture, SynTest, revenue growth management, data management, data visualization, managed services, customer loyalty, analytics academy, assortment optimization and integrated business forecasting.” analytic-edge.com ↗

On the record — “Demand Drivers is described as a six-step workflow (Input, Review, Modelling, Reporting, Simulation, Optimization) offered as full-service, SaaS, or SaaS plus FTE, ingesting data via API with manual UI load for ad hoc inputs. No validation method, holdout, backtest, stated assumption, limitation, defined refresh frequency or price appears on the page.” analytic-edge.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 — Three engagement models are named (full-service, SaaS, and SaaS plus FTE), and deliverables are identifiable: modelled cross-channel attribution, scenario simulation, budget optimization, and dashboards. Cadence is the weak half. The product page promises always-on insights aligned with planning cycles and frequent updates, without defining frequent anywhere I could find. The only concrete refresh I read is inside the Codeway case study, which describes weekly model results plus a secondary halo model. One instance in one case study is not a published standard, so this is not a stated cadence in the sense the rubric's top band asks for. That is between the two bands, which is why it scored Adequate.

On the record — “Demand Drivers is described as a six-step workflow (Input, Review, Modelling, Reporting, Simulation, Optimization) offered as full-service, SaaS, or SaaS plus FTE, ingesting data via API with manual UI load for ad hoc inputs. No validation method, holdout, backtest, stated assumption, limitation, defined refresh frequency or price appears on the page.” analytic-edge.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.

SaaS — software as a service: subscription software; as a client type it brings recurring revenue and metrics like churn and lifetime value.

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.

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. No fee, range, minimum, or licence term appears on the home page, the solutions page, the Demand Drivers page or the contact page. Three engagement structures are named but carry no numbers at all, so a buyer has no anchor and cannot tell whether they are entering a five-figure or seven-figure conversation. Also unresolved by anything published: whether the model licence, and access to the model itself, continues after an engagement ends, which matters more here than usual because two of the three engagement models are subscription-shaped. That is what the low band describes, which is why it scored Weak.

On the record — “No pricing, fee range, minimum engagement or licence term is published on the contact page, which lists thirteen office locations with a Singapore headquarters at 35 The Gateway West, 150 Beach Road, and names no leadership.” analytic-edge.com ↗

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

Stronger here: Nepa scores Strong on the same dimension.

Weak — 2 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 6 · Weak 10. The typical agency here scores Weak, and 6 of them score 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 — No independent client review base located. A G2 search returned HTTP 403 and could not be read first-hand, so no G2 rating or count is cited and absence there is explicitly not treated as a finding. No Clutch or Gartner Peer Insights entry for this firm surfaced. The site publishes one attributable client voice, the Codeway quote, and no reference programme. The corroborated third-party material that does exist is partnership and corporate news (Google Meridian certification, the Meta engagement, the C5i acquisition, MarTech Summit sponsorship), which is not client review evidence. Low weight is appropriate: measurement work of this size sells through procurement and referral, and a thin public review footprint is normal for the category.

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