How recommendations work

Where information comes from

Here's where each kind of content comes from, so you can weigh it accordingly:

• Calculated by SummaStack: match bands, recommendations, stack insights, “why it fits” notes and comparison scores. Calculated by SummaStack as described on this page.
• From accounting firms: reviews and ratings written by signed-in firm users. See the review policy.
• AI-generated: the short fit summaries on the Compare page, written by an AI model from the comparison scores and your firm's size and services.
• Product listings: descriptions, features, pricing and integrations come from vendor materials and SummaStack research. Anyone can report a listing that looks wrong.

There is no sponsored placement in SummaStack today. If that changes, sponsored content will be labeled “Sponsored” and kept out of recommendations and rankings.

How recommendations are chosen

SummaStack looks at the software your firm uses today and finds the categories with nothing in them. A tool marked “currently using” or “replacing” covers its category; one you're only evaluating, have planned, or used in the past doesn't. Categories you mark as not needed are skipped.

For each remaining gap, SummaStack shows a shortlist picked by CalCPA's editors, chosen for your firm's size where a size-specific list exists. If a category has no editor's list, SummaStack shows nothing for it rather than guessing.

Match bands

Each recommendation has a 0–100 score, shown as a band first. Until firms have rated an app, its score is just a starting value set by how it was picked, so no band is shown. Once ratings come in, the band appears, and Excellent match has to be earned with real evidence.

• Excellent match (80–100): A strong pick for firms like yours, backed by real ratings or peer usage.
• Strong match (65–79): Picked for firms your size, or a general pick with strong ratings.
• Potential fit (50–64): A good general pick for this category, not tailored to your firm's size yet.
• Explore carefully (0–49): Ratings from firms have pulled this pick below our usual bar. Compare before deciding.

Evidence tiers

Beside every score is a label saying how much real data backs it. A Curated 70 and an Established 70 are the same number, but the second is backed by real firms.

• Curated: No ratings or usage data yet, so the score is only its starting value.
• Emerging: A few real ratings from firms are starting to adjust it.
• Established: Backed by enough ratings or peer-usage data to be meaningful.

What moves a score

• The starting value.Every pick starts from a base. A pick chosen for your firm's size starts higher.
• Ratings.Reviews and the satisfaction ratings firms give tools in their stack move the score up or down. A handful of ratings barely moves it, so thin evidence can't swing a recommendation.
• Peer adoption. The top of the scale is reserved for evidence that firms your size actually use a product. Peer figures only switch on once a size group has at least 20 firms, so you never see a percentage based on a handful of firms.

Usage and rating figures are refreshed weekly.

What never affects a score

• Payment or sponsorship of any kind.
• A vendor's own reviews, ratings or stack entries for products they own.
• The “why it fits” and “things to consider” notes and editors' written reasons. They explain a pick; they don't change its score or order.
• Missing product data. If a field isn't known, SummaStack says nothing rather than guessing.

Stack insights

On My Stack, SummaStack points out categories with no tool, categories where your firm uses more than one tool, and tools in your stack that list integrations with each other. Insights come only from your stack and the product listings, and you can dismiss any of them.

Comparison scores

Compare scores each product on four areas: Quality & Reputation (ratings, reviews and service health), Capabilities (features and integrations), Value (pricing) and, when your firm has a profile, Firm Fit (your size and services). Where a product's data is missing, that area falls back to a middle value, so treat close totals as a tie.

Make your recommendations better

• Add every tool your firm uses to My Stack.
• Mark categories you don't need as not needed.
• Rate the tools you use. It's the most useful thing you can contribute.