ParseHub is one of the few serious visual scrapers still standing: you point at elements on screen, the tool infers the pattern, you get a file. If you're looking for an alternative, extraction quality probably isn't your problem. It's two lines on the pricing page that almost nobody reads before starting.
What ParseHub does better than most online tools
Three real strengths, to acknowledge before any criticism. First, it's a desktop application, which lets it interact with pages online scrapers give up on: forms, dropdowns, infinite scroll, content loaded after the fact. Second, visual selection handles nested structures — a list where each item opens a detail page to walk through — where many tools stop at a flat table. Third, the free plan isn't a crippled demo: 200 pages per run and five projects, no card required, is enough for a genuine trial.
Two free-plan clauses to read before you start
The first: on the free plan, all five projects are public. That isn't legal small print, it's an operational decision. A scraping project describes exactly what you target: which site, which pages, which fields. If your extraction reflects a commercial strategy — a sector, an area, a type of business — making it public amounts to publishing your prospecting plan. Private projects exist, they start on the paid plan at $189 per month.
The second: data retention is 14 days on the free plan. So you have to export on every pass. That only becomes awkward once extraction turns recurring — which is exactly when the free plan stops being a test bench and becomes a production constraint.
The one-week test protocol
Rather than comparing feature grids, run the test that settles it. Day 1: build the project on ten pages of your real target, not on a documentation example. Day 2: run it again unchanged and compare both exports row by row — a visual scraper that drifts shows it on the second run, not a month later. Day 3: count the fields still missing before you could place a call or send an email. Day 5: estimate how long it would take to source those fields elsewhere.
That last number is the only one that matters. It doesn't measure ParseHub's quality — which is good — but the distance between what the tool returns and what you actually need.
For French company data, what ParseHub doesn't cover
A general-purpose scraper returns what's displayed: a name, an address, a phone number, sometimes a website. To contact French companies you then need the link to the official register — SIRET, legal form, officers, active or struck-off status — and verified contact addresses. At outsend those steps sit in one chain: extracting an area from Google Maps, SIRET, VAT and RCS enrichment, legal data and company officers, deliverability checks. During the alpha, access is free, by application.
This isn't a ParseHub flaw: it does generic web scraping, and it never promised to know the French company register. It's a difference of scope, settled by looking at what you extract. Comparable trade-offs are collected in our comparison of prospecting tools; the case of a platform billed by compute is covered in the Apify alternative.
Matching grid
| Criterion | ParseHub | outsend |
|---|---|---|
| Nature | General-purpose visual scraper, desktop app | Chain dedicated to company lists |
| Sites covered | Any site, including interactive ones | Worldwide Google Maps + French Sirene register |
| Free plan projects | 5, public | Full alpha access by application |
| Data retention (free) | 14 days | File export, no expiry |
| Setup required | One project to build per site | An area and an activity |
| French legal matching | Not covered | Built in |
| Email verification | Not covered | Built in |
Two rows cut against us, plainly. If you need to extract a site that is neither Google Maps nor a company register — a catalogue, a listings site, an industry platform — ParseHub will do it and we won't. And if you want control over every extracted field, the visual control of a ParseHub project is finer than any tool that decides for you.
Questions asked before choosing
Does "public" mean my extracted data is visible?
The pricing page describes free-plan projects as public. A project holds the definition of the extraction: the target site, the pages walked, the fields kept. Treat that definition as exposed — reason enough on its own not to build a sensitive commercial target on the free plan.
Does ParseHub work on Google Maps?
Technically a visual scraper can interact with a map. But Google Maps isn't a paginated site: results load on scroll, depend on framing and area, and a project tuned for one city doesn't transfer to the next. That's the case where an area-native tool answers in one operation while a visual project needs tuning per target.
Is there an open-source alternative to ParseHub?
Free crawling libraries exist, but they don't replace visual selection: you write the code, you host the runs, you handle blocking. A real option if you have the skills, and a false plan if avoiding code was the point.
Is the free plan enough for occasional use?
For evaluation, yes. For production, the two clauses above decide: 200 pages per run covers a one-off extraction, 14-day retention forces you to export every time, and public projects settle it for anything touching a commercial strategy.
What happens when the site changes structure?
The project has to be reworked. That's true of every hand-configured scraper, ParseHub included, and it's the hidden cost of this class of tool: the time isn't in the first extraction, it's in the ones after.
When should you stay on ParseHub?
Varied sources, interactive sites, a need to control every field, and a team willing to maintain projects. It's a good tool for that work, and no specialised alternative will give you back that versatility.
And for a list of French companies to contact?
A dedicated chain is shorter: legal data and contact addresses are most of the work, and those are exactly the steps a general-purpose scraper leaves to you.