CDP Scoring Shouldn't Cost You a Fortune — Try AI Instead

Tuesday, 8:15 AM. A cold flat white on the desk, two monitors on, and the CDP 2026 Corporate Questionnaire open in one of them — 13 modules, a response window that closed last year in mid-September, and last year's B staring back at you from the scorecard PDF your CEO printed out and left on your chair.

You are an ESG manager at a mid-cap industrial listed in Paris. This is your third CDP cycle. The first year you answered maybe 60% of the questions and scored a C. The second year you hired a consultant, followed their advice, and climbed to a B. Now the investor relations team has been told by two asset managers that they "track CDP A List inclusion" and the ask, delivered in a one-line Teams message, is: push us to A- this year.

You re-read module 3 — Business Strategy — and remember that last year's response was written by someone who did not know your three-year capex plan. You scroll into module 6 — Emissions — and notice that the Scope 3 category 1 number there does not match the one in the ISSB S2 filing you published last month. Great.

There are three ways to deal with this. We tried all of them.

Option A: Hire a Sustainability Boutique (The "White-Glove" Route)

This is what most returning filers do once they realize CDP is not a weekend project. You call a sustainability boutique — ERM, Anthesis, South Pole, EcoAct, Sustainalytics' advisory arm, one of maybe a dozen credible names — and ask for a "CDP optimization" engagement. They have a standard SoW for exactly this.

Here's how it typically unfolds.

Week 1: Kick-off and scoring diagnostic. A senior consultant (the one you met in the sales meeting) walks you through your previous year's scorecard, points at the modules where you lost points, and proposes a scoring-uplift plan. You feel good. You sign.

Week 2-3: The senior consultant hands the engagement to a two-person analyst team you've never met. They send you a 47-item data request list. Half of it is stuff you already published in your CSRD/ESRS filing. The other half involves tracking down your facilities manager in Milan to confirm renewable electricity procurement percentages by site.

Week 4-5: Drafting. The analysts draft responses for all 13 modules. The quality varies in a way you don't notice until you read module 11 — Target setting — and realize the consultant has not reconciled your near-term target with the SBTi validation letter you received in January. You send it back. They redraft.

Week 6: Review call. The senior consultant re-appears, walks you through the final draft, flags three "risks" where they think CDP might deduct points, and suggests optional upgrades. You submit via the CDP portal. You pay the invoice.

Total time: 4 to 6 weeks of elapsed calendar, about 35 to 55 hours of your own time. Total cost: €25,000 to €60,000.

Scores land in December or January. If you moved from B to A-, the consultant takes the credit. If you landed at B again, the consultant explains that CDP tightened the bar this year. Either way, next April they call you back to renew.

What you get

• A polished, professionally written questionnaire
• A scoring diagnostic you could probably have written yourself
• Someone senior to sit next to your CSO when the board asks

What you don't get

• A response that reliably matches what you said in your ISSB, CSRD, and GHG Protocol reports
• Consistency year over year — your 2025 analyst is not your 2026 analyst
• Any knowledge transfer to your internal team

Option B: CDP Module on a Carbon or ESG Platform (The "Configured Workflow")

The second route, increasingly popular since CDP's 2024 shift to the unified Corporate Questionnaire, is to buy a platform that has a CDP module bolted onto its carbon accounting or ESG reporting core. The obvious names: Sweep, Position Green, Persefoni (which added a dedicated CDP workflow in 2024), Watershed, and Novisto. Each sells a configured questionnaire builder with pre-mapped fields, response libraries, and benchmark scoring.

The experience looks roughly like this.

Week 1: Implementation kick-off. Your customer success manager sets up the CDP module, loads your previous year's response as a baseline, and runs a gap analysis against the 2026 questionnaire. You learn that CDP changed 37 questions this cycle. Most are minor wording; four are substantive.

Week 2: Data integration. Your GHG inventory flows in from the platform's carbon module — if you were already using it. If not, someone on your team spends three days re-entering Scope 1/2/3 figures from PDFs. The platform pulls risk and opportunity narratives from a response library populated by a previous consulting engagement or by the platform vendor's in-house methodology team.

Week 3: Review and tuning. The platform flags responses that scored low last year and suggests improved language. Your team — or more realistically, one person on your team who has become "the CDP person" — reviews every suggestion, accepts some, rewrites others. The scoring simulator estimates an "A-" with confidence bands. You submit.

Week 4: Submission and audit prep. The platform archives evidence links and generates an internal audit pack.

Total time: 2 to 4 weeks. Total cost: €15,000 to €40,000 per year subscription, often bundled with a carbon module you are already paying for.

What you get

• A structured, repeatable workflow
• Scoring simulations that are directionally useful
• Response libraries that beat a blank page

What you don't get

• Automatic alignment with your ISSB S2, CSRD E1, and GHG Protocol inventory unless all of those also live on the same platform (they usually don't)
• A clean story for the 20% of questions that require narrative rather than numbers
• Protection from the "who configured this" problem — a sharp consultant using the platform gets an A-; a junior analyst using the same platform gets a B. Same tool, different output.

Option C: Formist AI Agent (The "One Platform, All Frameworks" Approach)

Formist, built by WeCarbon, is an AI-powered compliance platform that works like a knowledgeable colleague who has read every CDP scoring methodology memo since 2018 and also happens to remember what you said in your CSRD and ISSB filings. You talk to it, upload documents, and it drafts the questionnaire — module by module, with source citations back to the exhibits and datasets you've already produced for other frameworks.

Here's what that looks like for a returning filer aiming at A-.

Hour 1: You open Formist and say: "I'm filing CDP 2026 Corporate Questionnaire. We scored B last year. We want A-." Formist pulls your existing CSRD/ESRS disclosures, ISSB S2 filing, GHG Protocol inventory, and SBTi target letter — all of which already live in your Formist workspace because you used it for those — and builds a draft response for 11 of the 13 modules. It flags the two it can't answer yet: module 3's updated three-year capex narrative and module 11's newly required scenario analysis.

Hour 2-3: Formist walks you through the gaps conversationally. It asks you to upload the latest board-approved transition plan. You drop a PDF. It extracts the relevant figures, drafts a narrative, and asks you to confirm tone. You edit two sentences. It flags that your 2030 scope 1 target as filed in CDP last year no longer matches the SBTi-validated version. It proposes the corrected wording and shows a diff.

Hour 4-6: Scoring review. Formist runs its own scoring model against the CDP 2026 methodology, identifies eight places where the draft would lose points, and suggests fixes — things like missing board-level climate oversight evidence, incomplete Scope 3 category boundary disclosure, or a supplier engagement metric that CDP now weights higher. You accept six, rewrite two.

Hour 7-8: Cross-framework check. Formist runs a consistency pass: every number that appears in the CDP response is checked against your CSRD, ISSB, and GHG inventory. Two mismatches surface — both from rounding differences — and Formist proposes the version to keep. You approve. Export.

Total time: 1 to 2 working days. Total cost: per-use or flat subscription, typically a small fraction of a consultant engagement.

What you get

• A response that is internally consistent with every other framework you file
• The same scoring logic applied the same way in 2027 and 2028 — no personnel drift
• Full source-document traceability on every number

What you don't get

• A partner-branded PowerPoint to hand the board (though you can ask Formist to produce one)
• Somebody external to blame if CDP surprises you on scoring

The Six-Dimension Comparison

Here is how the three approaches stack up.

The column that matters most, and the one the CDP conversation usually skips: score consistency. Last year's B and this year's A- and next year's back-to-B is not really about CDP. It's about who drafted the response. Consultants rotate. Platform configurations get tweaked. The only way to hold the score steady is to hold the method steady.

What Happens as Your CDP Scope Grows

The three approaches diverge the moment CDP pulls more of your business into the questionnaire.

Boutique consultant: Costs scale roughly linearly with response depth. Add a second geography with its own Scope 3 complexity? Expect a 40% fee uplift. Add forest or water modules because CDP merged them into the corporate questionnaire and your sector now triggers them? Another line item. If your company grows through acquisition, the analyst team re-does discovery from scratch every year.

Platform module: Subscription is mostly flat, but configuration effort scales. Each new legal entity, each new supplier, each newly material topic means your in-house "CDP person" adds hours. And if that person leaves — ESG managers move roles roughly every 2.5 years — the platform is only as good as the handover document they wrote.

Formist AI agent: Effort scales sub-linearly. Uploading more documents is cheap. Formist learns your company once and re-uses the knowledge across frameworks and across years. The incremental Scope 3 category 11 disclosure for CDP reuses the same number you already calculated for ISSB S2, and Formist knows that, because it drafted both.

For a single-country mid-cap in year one, the three approaches feel more similar than different. For a pan-European group filing CDP Climate, Water, and (increasingly) Plastics modules while also producing a CSRD report and a TCFD-aligned annual report, the gap is enormous — and it compounds every cycle.

Why Waiting Until June Stops Working

The CDP calendar is not sympathetic to late starts. The 2026 cycle opened April 9 and the response window closes on September 16. Scores are published in December or early January. That's the public schedule. The less public schedule:

• CDP's scoring partners finish methodology updates in May — which is why a draft built in April often has to be revised in June
• Asset owners that rely on CDP data for portfolio decisions pull provisional responses in October, weeks before final scores are published; if your submission is late or weak, you miss that window entirely
• The new 2026 sector-specific modules (expanded for financial services, apparel, and construction) add response depth that typical 4-week engagements are scoped too tightly to absorb
• Assurance expectations are climbing. More investors now expect limited assurance on at least the emissions portions of CDP responses, and assurance providers need finished drafts by late July to be realistic

If you start the consulting scoping conversation in June, you are already choosing between paying a premium or filing incomplete. The A List is awarded to roughly 15% of disclosers globally — around 400 companies out of more than 24,000 — and the gap between A- and A is almost always data completeness and cross-framework consistency, not narrative brilliance.

A Formist AI agent that drafts the first pass in a day — in April, not July — leaves you the rest of the cycle to focus on the two or three narrative modules that genuinely need human judgment: transition planning, board oversight, the specifics of your TCFD scenario analysis. That's where the scoring points you actually need are hiding.

The Practical Path Forward

You don't have to fire your boutique consultant to try this. A sensible sequence:

1. Run your 2026 draft through a Formist AI agent first. Upload last year's CDP response, your CSRD filing, your ISSB S2 report, and your GHG inventory. See the draft it produces in a day. Note the cross-framework mismatches it flags — those are real, and you had them last year too.
2. Use the Formist output as the brief for a narrower consulting engagement. Instead of paying €40K for soup-to-nuts drafting, pay €8K–€15K for a senior consultant to review the narrative modules — strategy, governance, scenario analysis — where human judgment earns its keep.
3. Commit to one method across cycles. The reason your score drifts is that your method drifts. Pick one engine for the mechanical work — ideally one that also handles your CSRD, ISSB, and GHG filings — and hold it steady.

The A List is not really a drafting problem. It's a consistency problem. The companies that stay on it are not the ones with the best consultants. They're the ones who answer the same questions the same way, year after year, with the same numbers tied to the same sources. That's the part software does well. The narrative, the judgment, the board conversation — that's yours. Stop paying consultants to do the first part. Start paying them, sparingly, for the second.

Formist is an AI-powered compliance platform built by WeCarbon, a climate tech company with offices in Paris, Shanghai, and Dubai. It supports CDP, CSRD/ESRS, ISSB, GHG Protocol, CBAM, EU Taxonomy, SBTi, and 15+ other sustainability frameworks under a single workspace.