Webshare: the quiet mid-tier contender

Four years of scraping has taught me to distrust proxy marketing. SERPwatch ran 2M rank checks a month on a startup budget, and the proxies went unreliable right after the AppSumo customers onboarded. So when I put six providers through the same 20-site probe set in September 2026, same transport stack, same c1/c4/c8 ramp, same 2,380 requests, I wasn't hunting the fastest node or the cheapest gigabyte. I was hunting the one that lets you start at $3.50.

Webshare is the first proxy company I've seen openly market itself as SEO proxies, so I bought their 1GB rotating-residential plan, no minimums, no upsells, and ran it through the same harness as everyone else.

The results came back honest. 71.7% success across 20 real sites, a 0.5% timeout rate, and a p50 of 1,242ms that barely moved from c1 to c8. 150 unique IPs out of 150. And a pool of 51 ASNs across 30 countries, half what the top providers offer. The most accessible entry into residential proxies, with the best support bot in the industry. The quiet mid-tier contender.

• $3.50/GB is the lowest entry price in the market: no other provider in my test lets you start at 1GB
• 71.7% success on real sites: mid-pack, with the third-lowest timeout rate (0.5%)
• 150 unique IPs out of 150 (100% rotation): no repeats across the run
• Webby, the AI support bot, told me rotating residential beats their own marketed datacenter IPs for SERP
• Pool diversity: 51 ASNs, 30 countries: half the market average, and it shows on PerimeterX walls
• No scraping API, no HTTPS: raw endpoints only (http, endpoint:port, socks5)

The 20-site sweep that held its shape

The probe (c1/c4/c8, 10 sessions per level, no retries) came back clean: 65.0% success at c1, 66.2% at c4, 75.5% at c8. Success climbs as concurrency ramps, the same shape as oxylabs and decodo.

The timeout rate was 0.5%, about 12 hanging requests out of 2,380, behind only decodo (0.1%) and oxylabs (0.3%), and miles ahead of byteful's 1.4% or shifter's 4.5%.

The auto-tier story holds too. With the browser tier (camoufox then playwright-stealth), target clears at 90%, second-best of the six and a hair behind oxylabs at 91%, and reddit at 99%. Amazon, Booking, Agoda, IMDB, and OzBargain all sit at 98-100%. The hard walls are the universal ones: slickdeals and tripadvisor at 0%, same as every provider in the test.

Where the mid-pack shows is the third-party praise versus my numbers. Webshare's marketing resellers quote a 99%+ infrastructure success rate; I measured 71.7% on 20 real sites. That's the difference between a CDN ping test and the sites you actually care about.

The latency that doesn't flinch

Webshare's p50 of 1,242ms isn't the fastest. Byteful leads at 961ms, decodo at 1,191ms. But ramp it from c1 to c8 and it barely moves: 1,181ms at c1, 1,162ms at c4, 1,245ms at c8.

Flat and predictable, which at 2-5M scrapes a month beats raw speed. A 100ms spike across 100 sessions cascades into retry logic stacked on retry logic.

The European gateways are the sleeper win: de 567ms, fr 604ms, gb 590ms. That's the fastest set of EU gateways of all six providers I tested. If your targets are Europe-heavy, that's a real edge.

The US gateway sits mid-pack at 1,002ms (byteful 750ms, oxylabs 1,265ms), and the global pool is 1,024ms across 30 countries. It's not a global-speed product; it's a regional one done well.

The pool that's honest about being small

Rotation is flawless: 150 requests, 150 unique IPs, 100% rotation. No repeats, no drift. But the pool itself is the smallest I measured: 51 ASNs and 30 countries, against shifter's 117, oxylabs' 102, and decodo's 94.

On the default gateway the egress is UA-heavy (Ukraine 10%, Turkey 8%, France 8%, Romania 8%, Netherlands 6%), not US-heavy at all. The geo gateways respect their tags, us and de at 100% egress match, but only if you set them.

That's where the pool reputation bit me. Zillow, a PerimeterX-fortified site, came back 0% with the same browser tier and sites, while decodo cleared it at 73%, dataimpulse at 95%, and shifter at 96%. Same rotation, same transport stack, different ASN reputation.

That's the honest read on Webshare's pool. What you lose in diversity, you lose on the reputation-sensitive walls. For the fast tier and the mid-tier Cloudflare sites, it's a solid performer. For the hardened ones, the pool is the ceiling.

The dashboard that feels like a builder's workspace

Most proxy dashboards are billing portals with endpoints tacked on. Webshare's is a workshop. You don't buy a plan; you build one. Choose 1 to 25,000 IPs, bandwidth from 250GB up, country, state, city, even ZIP code, threads, network priority, and IP refresh frequency, and the base price doesn't move. It's the build-your-gaming-PC vibe, and it's genuinely the most customizable plan flow I've seen in the six I tested.

The endpoint generator is where it shines. It asks for ASN, location, and sticky-or-rotating. No digging your account password out of the environment like oxylabs forces.

Sticky sessions are keyed on the username, not the port, so you can hold multiple sessions at once (jidihkrw-674378, jidihkrw-674379). That detail matters when you're running multi-provider scrapes. Protocols are http, endpoint:port, and socks5. No HTTPS, same as the rest of them.

The catch is the express checkout. I picked the 100-proxy datacenter package and it jumped straight to the payment page, one step from subscribing, with no way to customize to dedicated or private datacenter IPs. I had to hunt down the choose-plan page myself. A product this configurable shouldn't hide its configurator behind a checkout funnel.

The support bot that talks straight

Webby, Webshare's AI chat agent, is the best support chatbot I've tested in this niche, better than a lot of real human support.

Here's the test that won me over. Webshare's landing page markets datacenter IPs as their SEO solution. I asked Webby whether rotating residential or datacenter IPs are better for SERP scraping. It said rotating residential, and told me its own landing page was misleading. A support bot that openly corrects its company's marketing is a trust signal I've never seen before.

It also confirmed the 100-proxy datacenter package can be customized (dedicated, private, or shared), a detail the checkout flow never showed me, and pointed me to the plan builder when I couldn't customize.

For the indie hacker who's been burned by bots that parrot sales copy, that's the difference between support and a sales funnel. I never needed to escalate to a human.

The $3.50 entry that changes the math

The industry norm is a $2/GB headline hidden behind a 10GB minimum. Oxylabs makes you buy $30 of traffic, decodo $11.25 for 3GB. Webshare sells 1GB for $3.50. No minimums. That's the lowest entry price in the market, and it's a testing ground, not a reference floor.

The catch: it's a testing ground, not a scale product. My plan capped out at c8 concurrency. Dataimpulse is the only provider in my test clean through c16. The free tier is 10 datacenter IPs with 1GB/month, a reference floor, not a service.

And the customization costs are honest but hidden behind upsells. 500 concurrency by default, 3,000 with an add-on, and a "high priority network speed and stability" add-on that made me wonder whether the network degrades without it.

So you trade cost for control. That's the Webshare game, and for a builder testing a new use case, $3.50 to find out is the right hook.

The only thing missing

Webshare has no web scraping API. What exists is an account-management API: registration, login, sub-users, traffic limits, proxy configs. Not a scraping endpoint. No browser rendering, no HTTPS, no prebuilt SERP or review extractors. Just raw endpoints: http, endpoint:port, socks5.

That means you're on your own with automation. My coupon system runs Playwright, Puppeteer, Selenium, and BS4 anyway, so raw endpoints suit me. For teams that want a one-click scraper, this is a wall, and the full-stack alternatives (ScrapingBee, Zyte) cost more.

SERP tracking: not tested yet

I haven't run a SERP benchmark yet, so I have no measured numbers for the thing Webshare markets itself on.

What I have: Webshare is the first provider I've seen openly position itself as SEO proxies, its own bot recommends rotating residential over its marketed datacenter IPs for SERP, and the dashboard's country/state/city/ZIP targeting is exactly what local SERP work needs. I'll run the SERP test and update this section.

The verdict: the quiet mid-tier contender

Webshare is the most accessible entry into residential proxies: $3.50 to start, the best support bot in the industry, reliable mid-tier transport that holds its shape under load. The catch: the pool is half the diversity of the top providers, and it shows on reputation-sensitive walls like zillow.

Alternatives

If Webshare's pool diversity or c8 ceiling doesn't fit, the honest alternatives from the same 20-site test set:

• Decodo: 76.9% success, 0.1% timeouts, $0.63/1k. The most reliable transport I tested, at a mid-tier price. The entry is 3GB for $11.25.
• Dataimpulse: $0.15/1k, the only provider clean through c16. The budget-and-concurrency pick. You pay in latency (2,114ms) and a 2.8% timeout rate.
• Shifter: $0.33/1k and the most diverse pool (117 ASNs), but the worst reliability I measured, 4.5% timeouts. Only if you can babysit runs.
• Oxylabs: prepackaged all-in-one at $30/6GB min. For teams that don't want to customize, if you can afford the entry.

The interesting question is whether the pool catches up to the product. Webshare's transport and support are good enough to be a primary provider; the pool is what keeps it mid-tier. I'll keep testing as the SERP data lands, and update this review.