The SEO tools our guests swore by, and what replaced them

About half an hour into part two of my catch up with Mike King, we got onto how much had changed since our first chat. I said Followerwonk was "no longer around". Then, about ten seconds later, I said it still existed and Moz was trying to sell it. Then I said I thought they'd already sold it.

Mike, very fairly, called it: "It's good to know it's still around. Yeah, let's fact-check that one."

So I have. Nine years late, and not just for Followerwonk. I went back through every WojCast transcript, including the interviews I recorded after Season 1, and pulled out every tool a guest leaned on. Then I checked what happened to each one.

Most of the tools are dead, sold or renamed. I still need every job they did.

The scorecard

Finding people

Three of the dead tools did the same job. They helped you find the humans.

Mike used Followerwonk to turn a persona into real people: "I can essentially search for those people in Followerwonk... Followerwonk is kind of like the best because it gives you Twitter right there." It was the most useful thing Moz owned that wasn't about search, which is exactly why it went. Rand told me in 2016 that Moz was "returning to its SEO roots and going to be focused on SEO for the foreseeable future." That refocus is when Followerwonk was cut loose. So for the record: it was still around, Moz was selling it, and the sale went through in 2017. Two out of three for me, in ten seconds.

When I interviewed Tim Soulo in late 2018, he explained how he used to build guest post lists before Ahrefs. He'd Google "social media blogs you should read" and then, "I think I was using Alexa at the time to see which of these blogs were getting more traffic... But right now I would totally use Ahrefs for this" (the written interview is here). Tim works at Ahrefs, so that last bit may not be entirely neutral. It's still what most people do.

The one I miss is Question Hub. When I spoke to Lily Ray in 2022 she described it as "questions that people didn't find an answer to on Google. So they like specifically submitted their question to Google. And like, you get some really emotional stuff in there" (more in the written interview). A few months later Google closed it. Search Engine Land wondered at the time whether the idea would turn up in Search Console instead.

The job hasn't moved, though. Lily's whole point was about understanding where the searcher is coming from, which is understand the context. You can still get there. Read your site search logs, your support inbox and the Reddit threads in your niche. They're the same unanswered questions, just less tidy.

Links and crawling: the survivors

Funnily enough, the tools that did the grunt work mostly survived.

Jon Cooper's complaint about Open Site Explorer in 2014 was blunt: "the size of their index is pretty poor compared to the other two." Moz clearly agreed. When they replaced it with Link Explorer, the headline was an index more than 20 times bigger. Rand had already told me Moz was using machine learning in Open Site Explorer for its Domain Authority and spam scores, and that thinking carried straight over.

Paddy's kit in 2013 was Screaming Frog, "aside from Excel, which kind of goes without saying Excel's brilliant", SEO Tools for Excel (which annoyed him because it didn't work on a Mac) and BuzzStream. Screaming Frog is still shipping major releases. SEO Tools for Excel is still around. And his favourite free tool, Google's advanced search operators, still beats "spending loads of money on tools" for finding prospects. We wrote up the operators we use most years ago and most still work.

When I asked Mike what was left of his old kit, we counted three tools: Screaming Frog, Followerwonk and Majestic. Screaming Frog and Majestic were still going.

Google gives, Google renames

Google's own tools were the least stable of the lot.

Gary told me in 2016, "I think AMP becomes more and more important for us." He had a good reason. His favourite news site took about twenty seconds to load in Australia. Three years later Paul Shapiro was experimenting with AMP stories and said, a bit prophetically, "Google's certainly not known for keeping some of their technology around" (Paul's interview). In 2021 AMP stopped being the ticket into Top Stories. The job, fast pages on slow phones, is now just called making a decent website.

When I spoke to Jeff Sauer in 2022 he had the Universal Analytics dates down cold: "June 30th, 2023 is the last day that it will collect data" (Jeff's interview). He was right. He also warned that anyone without GA4 running by the end of 2022 would lose their year on year comparisons, and a lot of businesses did.

Then there's Data Studio. Paul's view on reporting was simple. Manually pulling data into Excel "is a waste of time, in my opinion. You got Data Studio, and Data Studio is fantastic." Google renamed it Looker Studio in 2022. In April 2026 they renamed it back. If you built your agency reporting on it, you've now updated your proposals twice to end up where you started.

So use Google's free tools, but keep a copy of your data somewhere Google can't rename.

The no code machine learning kit

Mike was showing marketers how to do machine learning without code in 2017, which felt a bit out there at the time. His kit was MonkeyLearn and Orange, plus API.ai hooked up to Zapier to answer the five standard objections people send back to outreach emails. His pitch was that none of it needed maths: "it's really just like everything else we do in marketing and digital marketing, it's guess and check."

API.ai became Dialogflow within months. MonkeyLearn went to Medallia. Orange is still free and open source, and so is KNIME, which Paul's team used to "cut out engineers" on data work. The job itself, getting a machine to sort, label and reply, is now something anyone does in a chat window. Mike was early, not wrong.

Digg is the cautionary tale. Oli Gardner ran through the social channels from his early days, Digg, Delicious and StumbleUpon, and shrugged: "technology's come and go." Digg came back in 2025 and shut its app again in March 2026. Digg told TechCrunch that when the beta launched "we immediately noticed posts from SEO spammers". We did that, as an industry.

What I'd do with this

• List your tools by the job they do, not by name. Crawl, find people, report, test. When one dies, you know what you're replacing.
• Export your data out of anything free. Analytics, reports, link lists. Paul's Wayback Machine trick, comparing the archived version of a site with the live one to find lost pages, only works because someone kept a copy.
• Keep one manual method for every automated one. Paddy's search operators, Lily's forum reading. They never get sunset.
• Treat every new AI tool like Followerwonk in 2012. Use it hard, and assume it might be gone, sold or renamed in four years.

None of the dead tools ever told anyone what their customers felt. They just made it quicker to go and find out. It's the same idea as you can't automate empathy, which has outlasted everything on this list. Mike's full episode on personas and data is still one of my favourites.