The method

Take values seriously. Score honestly. Show your work.

A transparent, evidence-based screening tool for advisors whose clients actually mean it. No committee in Boston deciding what's good. No pre-packaged ideology. A values vector the client builds, a classification trail the advisor can verify, and a report that prices the conviction in basis points.

Thesis

Your client's values are theirs, not a committee's.

Traditional ESG scoring fails at something fundamental: it assumes a single definition of “good” applies to everyone. Your clients don't share one worldview, and an honest tool doesn't pretend they do.

Allocate Good starts from the other direction. We don't score companies in the abstract. We score them relative to the values the client just told you they hold. The same company can be a perfect fit for one household and a hard exclusion for another — and that's the point. From there the platform builds the scorecard, the reasoning, and the cost of conviction, and hands them to you to deliver.

Four steps, in order

Profile the client

The client takes a nine-minute survey that weights each of nine value areas, then rates the specific industries inside the ones they care about.

The output is a values vector — 65 category weights between −1 and +1 that describe what the client wants their money to steer toward and what they want it to avoid. A vegan and a cattle rancher, a pacifist and a defense hawk — each produces a different vector.

The platform has no opinion about which weights are correct. It only asks the questions clearly enough that the client can answer honestly.

Classify each company

Every company is read against all 65 categories using three independent sources in sequence — annual filing, reference material, and current reporting — not headlines or sentiment.

Stage one reads the company's 10-K (or foreign-filer equivalent). Stage two reads the Wikipedia article. Stage three runs targeted web searches per category. Each stage independently classifies all 65 categories; disagreements are resolved by weighted synthesis.

Every classification links back to the source paragraph. No category is ever asserted without a traceable citation.

Score the portfolio

Each holding's classification is scored against the client's vector, then rolled up to a portfolio-level alignment score from 0 to 100.

The score uses a dampened denominator — categories the client doesn't care about don't dilute categories they do. Uninvolved categories contribute neutral. Hard exclusions override everything else.

The same score algorithm runs on funds as on individual tickers, so a client's existing ETF allocation can be evaluated the same way as a direct-indexed portfolio.

Price the conviction

The report tells the client what holding their values-aligned portfolio costs, measured in basis points of expected return versus VOO over the trailing twelve months.

Values-based investing that doesn't price itself is just marketing. A fee-only advisor should be able to say: 'holding this portfolio instead of VOO costs you forty-two basis points a year. Here is why. Here is what you get for it.'

The PDF is prepared under the advisor's lockup. The client never sees a marketing brochure from Allocate Good — they see a report from their advisor.

The deliverable

A PDF you hand across the table.

Page 1

Scorecard

Alignment score, cost of conviction, the client's three largest misalignments. One page. Set in the advisor's lockup.

Pages 2–4

Reasoning

A research-analyst paragraph per position that drives the score up or down. Every claim cites a specific paragraph of a filing, a Wikipedia article, or a web source.

Page 5

Trade proposal

Tickers, weights, and the expected-return delta versus the benchmark. What it costs. What it gets you. Signed over to the advisor for execution.

Evidence

Every claim is traceable. Every classification is reviewable.

Classifications are cached, versioned, and link back to the exact source paragraphs they were built from. Low-confidence categories are flagged for human review. Human corrections are injected back into the pipeline on the next run, so the system gets more accurate as more advisors use it.

You can open any holding on a client's scorecard and see the citations that produced its score — not a methodology PDF, the actual sources. If a classification is wrong, you can flag it and a reviewer will look at it.

Ready to try it

Request access. Bring one household. Send them the survey and score the portfolio they're already in.

Request access Back to home →