What Dogpile is, and where its results come from
Dogpile is a metasearch engine. You type one query; it sends that query to several other search engines' result feeds, merges what comes back, removes duplicates and shows you a single list. It has web, image, video, news and maps tabs, a dog mascot called Arfie, and a homepage tagline that reads "Searching the web since 1996."
The defining fact about it, and the one most readers get wrong, is that Dogpile has never run a web crawler — the automated program an engine uses to fetch pages and build an index (its own stored copy of the web). There is no Dogpile bot, no Dogpile user-agent, no published crawl policy and no Dogpile index. That is not a decline from some earlier state. It is what metasearch is: a layer on top of other people's indexes.
That makes the verdict at the top of this page a neutral structural description rather than a criticism. Dogpile resells other engines' results because reselling other engines' results is the product.
What has changed is the context around it. Dogpile today is operated by InfoSpace Holdings LLC, a subsidiary of System1, Inc., an advertising company in Marina del Rey, California. Its footer, checked on 19 August 2026, reads "© 2026 Infospace Holdings LLC, A System1 Company." So the honest description of the site in 2026 is a monetised search portal owned by an advertising business, carrying a well-known 1990s brand.
Whose results you are actually seeing
Dogpile does not tell you on its homepage. As of 19 August 2026 neither the homepage nor the footer names a single supplier; you have to go looking for the About page.
When you find it, the About page says Dogpile "returns all the best results from leading search engines including Google and Yahoo!", and Dogpile's app listing also names Bing. Wikipedia additionally lists Yandex. Two large caveats belong with those names. First, the About copy is undated and reads as legacy text: Yahoo! search results have themselves been Bing-powered since 2010, so naming Google, Yahoo! and Bing may describe two upstream engines rather than three. Second, System1 does not publish its syndication agreements, so the weighting between suppliers — and whether Google is still a live feed at all — is not documented anywhere. The accurate formulation is that Dogpile claims Google and Yahoo!; the current mix is not disclosed.
Historically the list was longer and more interesting. At launch and through the late 1990s Dogpile queried Yahoo!, Lycos, Excite, WebCrawler, Infoseek, AltaVista, HotBot, WhatUseek and the World Wide Web Worm — a roll call of an era when no single engine covered much of the web.
Dogpile also used to label each result with the engine or engines it came from. Whether that attribution still appears on every result could not be confirmed for this page: a live results page fetched on 19 August 2026 returned only a JavaScript loading shell, because results are rendered in the browser. Anyone relying on provenance should check it in a browser rather than take it on trust.
The case for metasearch, and the study Dogpile paid for
Metasearch has one genuine argument in its favour, and Dogpile made it more forcefully than anyone. In 2005, research commissioned by Dogpile and carried out with researchers at Queensland University of Technology and Penn State found that only about 1.1% of first-page results overlapped across Google, Yahoo! and Ask Jeeves simultaneously, and 3.2% across any two of them. If the leading engines agree on almost nothing at the top of page one, seeing several at once has obvious value.
Two qualifications. The study was funded by Dogpile — it is real research, but it is not disinterested research, and it should always be cited with that attached. And the finding is narrower than the marketing built on it. Low overlap shows the engines differ; it does not show that a merged top slice is bigger. A metasearch page is a deduplicated sample of several engines' first pages, which is typically far fewer results than any one of those engines would have given you on its own.
Dogpile was, for a while, a genuinely significant destination on the back of that pitch. It won J.D. Power and Associates awards for best Residential Online Search Engine Service in 2006 and 2007, and its "Search and Rescue" charity programme reported passing $100,000 raised for animal charities in December 2008.
Who owns it: Go2Net, InfoSpace, Blucora, OpenMail, System1
The ownership chain is the story of what happened to the whole 1990s metasearch generation.
- November 1996 — launched by Aaron Flin as an independent metasearch site.
- August 1999 — acquired by Go2Net, Inc., which already owned MetaCrawler. The two best-known metasearch brands are now under one roof.
- July 2000 — InfoSpace, Inc. acquires Go2Net for around $4 billion in stock, at the very top of the dot-com market. The deal is a textbook case of bubble-era overpayment: InfoSpace's market capitalisation fell from roughly $31 billion in April 2000 to a fraction of that within two years.
- June 2012 — InfoSpace, Inc. renames itself Blucora after acquiring the tax-software company TaxAct. The search properties become a segment of a diversified company rather than its identity.
- July 2016 — Blucora sells the InfoSpace search business, including Dogpile, WebCrawler and MetaCrawler, to OpenMail LLC for $45 million cash. Sixteen years after a $4 billion purchase, that number is the clearest available measure of what the brands were worth.
- 2017 — OpenMail renames itself System1.
- January 2022 — System1 completes a merger with the Trebia Acquisition Corp. blank-cheque vehicle and lists on the New York Stock Exchange as SST.
Watch two common errors here. Dogpile is not still owned by the 1990s InfoSpace: that company became Blucora in 2012 and sold the search business in 2016, and "InfoSpace Holdings LLC" is a System1 subsidiary that inherited the name. And Dogpile, WebCrawler and MetaCrawler are not rivals — they have shared an owner since 2000 and share infrastructure today.
How it makes money
Advertising. Dogpile earns from paid search listings and display advertising placed on and around its results, plus commerce placements — the homepage as checked on 19 August 2026 carried merchant shortcuts to Amazon, Walmart and Lowe's. There is no subscription tier and no paid API for end users; developer.dogpile.com survives as a legacy developer and affiliate resource.
Parent System1 describes its business publicly as responsive acquisition marketing: buying traffic cheaply and monetising it through advertising and search-syndication revenue share, across an owned-and-operated portfolio of search and content websites. Dogpile is one of those properties.
Why this matters to a user rather than an investor: an engine that owns an index competes on the quality of that index and has a reason to describe it. An arbitrage property competes on the yield of the results page. Dogpile does not publish its feed sources or the ratio of paid to organic listings, and nothing in its business model gives it a reason to start.
Privacy and data handling
Dogpile makes no meaningful privacy claim and does not market itself as a privacy engine. Its Privacy Policy and Terms of Use link out to System1's legal pages — confirmed on the homepage on 19 August 2026 — which means Dogpile users are governed by an advertising-technology company's general privacy policy rather than a search-specific one. The site is under United States jurisdiction, where there is no comprehensive federal privacy law.
There is also a structural point that applies to every metasearch engine, including the ones that do make privacy claims: your query is transmitted to third-party engines by definition. Metasearch cannot offer end-to-end query privacy unless the operator proxies the request and strips identifiers, and Dogpile makes no such claim.
In November 2019 Dogpile was named, along with WebCrawler and MetaCrawler, in commentary about data flowing into System1's advertising infrastructure, raised at the time in the context of System1's investment in Startpage. That is advocacy writing, not a regulator's finding, and the framing deserves care. The corporate facts under it are documented and undisputed: System1 owns all three brands, and System1's Privacy One Group acquired a majority stake in Startpage in October 2019. No regulatory enforcement action against Dogpile specifically was found.
Who it suits, and who it does not
What it is good at. The one thing metasearch is structurally good at: showing you that different engines return different things. If you genuinely want a quick second opinion without visiting three sites, Dogpile does that, and its interface is fast and uncluttered by the standards of modern search results pages. Thirty years of brand continuity also make it interesting in its own right for anyone studying how the consumer web aged.
What it is bad at. Everything that depends on owning an index. It cannot rank better than its suppliers, because it has nothing to rank with. It cannot promise freshness. It cannot reliably support advanced operators, because operator support depends on the upstream engine. It cannot be audited: you have no way of knowing whose results you are looking at or how heavily paid listings are mixed in. It offers no privacy benefit — the query goes to an advertising company and then onward to its suppliers. It has no AI answer layer; none has been announced as of August 2026.
Who should look elsewhere. Anyone choosing an engine for privacy. Anyone who needs an index independent of Google and Bing — that is Mojeek or Brave, and it is not Dogpile. Anyone doing serious research who has to be able to say where a result came from. Dogpile is best understood as a legacy consumer portal, not a research tool.
Common misconceptions
"Dogpile searches the whole web itself." It has never had a crawler. It has always been a front end over other engines' results, from 1996 to today.
"Metasearch gives you more results than Google." It gives you a merged, deduplicated view of the top slice of several engines — usually far fewer total results than any one of them. The 2005 overlap study shows the top results differ; it does not show Dogpile's coverage is broader.
"Dogpile is dead." It was live and returning results on 19 August 2026. This is where it differs sharply from Mamma, the other well-known 1990s metasearch brand, whose site is live but no longer has a search function at all.
"It is a privacy alternative." It is owned by an advertising company and governed by that company's privacy policy. There is no anonymity claim anywhere on the site.
"It is still owned by InfoSpace, the 1990s dot-com." Same brand string, different company — see the ownership chain above.
"Dogpile, WebCrawler and MetaCrawler are competitors." They have been sister properties since 2000 and are all System1 properties today.
Frequently asked questions
Does Dogpile have its own search index?
No. Dogpile has never operated a web crawler and has never held an index of its own. It is a metasearch engine: it sends your query to other engines' result feeds, merges what comes back and deduplicates it. There is no Dogpile bot and no Dogpile user-agent, because there is nothing for one to do.
Which search engines does Dogpile use?
Dogpile's own About page names Google and Yahoo!, and its app listing also names Bing; Wikipedia adds Yandex. None of this is dated, and System1 does not publish its syndication contracts, so the current mix and the weighting between suppliers are undisclosed. Note too that Yahoo! results have been Bing-powered since 2010, so that list may describe fewer upstream engines than it appears to.
Who owns Dogpile?
InfoSpace Holdings LLC, a subsidiary of System1, Inc., an advertising company in Marina del Rey, California. Dogpile was independent from 1996, went to Go2Net in August 1999, to InfoSpace in July 2000, stayed with Blucora after its June 2012 rename, and was sold to OpenMail in July 2016 for $45 million. OpenMail became System1 in 2017.
Is Dogpile still working in 2026?
Yes. The homepage loaded and returned web, image, video, news and maps results when checked on 19 August 2026, and the footer carried a 2026 System1 copyright. That distinguishes it from Mamma, a contemporary metasearch brand whose website is commercially active but no longer offers search at all.
Is Dogpile a private search engine?
No. Dogpile makes no privacy claim, publishes no search-specific privacy policy, and links its Privacy and Terms pages straight to System1's. Beyond that, metasearch is structurally poor for privacy: the query is forwarded to third-party engines by design. Anyone choosing on privacy grounds should look at engines that make and document an explicit commitment.
Is Dogpile better than Google?
Different, not better. Dogpile cannot outrank its own suppliers because it has no index to rank with; what it offers is a merged view of several engines' top results. The 2005 study it commissioned found little overlap between the leading engines' first pages, which is a real argument for a second opinion, but it is an argument about variety, not about depth or accuracy.
Are Dogpile, WebCrawler and MetaCrawler the same thing?
They are three brands owned by the same company. All are System1 properties run through InfoSpace Holdings LLC, and they have shared an owner since Go2Net and then InfoSpace consolidated them in 1999 and 2000. The interfaces differ; the ownership, the business model and the practice of buying results rather than crawling do not.