How to Request Custom Data (Bulk G2, Capterra, Upwork and More)

What you can request

Any dataset we hold is fair game. The most common requests are bulk pulls that would take hundreds of clicks to assemble through the UI: every G2 product in a category with its review counts and ratings, or every Capterra listing in a vertical with the full review text attached.

• G2: products, categories, ratings, review text and competitive insights
• Capterra: listings, reviews, feature gaps and category analysis
• Upwork: job posts, pain points and category demand signals
• App Store and Play Store: apps, ratings and review text
• Reddit: pain points, solution requests and derived SaaS ideas
• Product Hunt: launches, upvotes and maker data
• Stripe Index: companies taking payments on Stripe, by category and business model
• Funded companies, acquisition listings and TrustMRR revenue data

How to write a request we can fulfil fast

The single biggest factor in turnaround time is specificity. A request that names the dataset, the filter, the columns and the time window can often be pulled the same day. A request that says 'send me G2 data' comes back to you as a question, which costs both of us a round trip.

Write the request the way you would write a spec for the file you want to open. Name the rows, name the columns, name the cutoff.

• Name the exact category or keyword, not the broad area
• Say which columns matter, and which you do not need
• Give a date range if recency matters to your analysis
• Set a floor if it helps: minimum review count, minimum rating, minimum MRR
• Tell us the use case, since we can often suggest a better cut than the one you asked for

A good request versus a vague one

Compare these two. The first arrives as a file. The second arrives as a follow-up question.

• Good: 'Every G2 product in Project Management with 50+ reviews. Columns: product name, vendor, category, review count, average rating, pricing tier, and the full text of reviews posted since January 2025. CSV.'
• Good: 'All Capterra listings in HR software, with feature gap analysis and the ten most recent reviews per listing. Excel, one sheet per listing is fine.'
• Vague: 'Bulk G2 data.'
• Vague: 'Everything you have on SaaS tools.'

What happens after you submit

Every request moves through a status lifecycle, and you get an email at each step. You can also see the full history of any request on the Data Requests page.

• Pending: submitted, waiting for us to pick it up. You can still cancel at this stage
• In review: we are scoping what it takes to pull
• In progress: we are running the query
• Needs info: we need more detail. Reply on the request and it goes back into review
• Fulfilled: the file is ready, with a Download button on the request and a link in your email
• Declined: we cannot fulfil it, with a note explaining why

Limits and turnaround

You can have five open requests at once. Fulfilled, declined and cancelled requests do not count against that, so the limit only ever throttles work that is actively queued. Turnaround depends on the size and the complexity of the cut. A well-specified pull from a single dataset is usually fast. A cross-source join with derived analysis takes longer.

Download links occasionally carry an expiry date, shown on the request and in the email. Pull the file down when you get it rather than relying on the link staying live indefinitely.

FAQ