RankScience closes $1.8M seed — and now only wants to replace human SEO staff if you don’t have any

A couple of years ago YC-backed RankScience, which offers AI-enhanced SEO split-testing, put a few SEO experts’ noses out of joint when the fledgling startup brashly talked about replacing human expertise with automation.

Two years on its pitch has mellowed, with the team saying their self-service platform is “augmenting human SEO ability rather than replacing them”.

The startup has also — finally — closed a seed round, announcing $1.8M led by Initialized Capital, along with Adam D’Angelo, Michael Seibel, BoxGroup, Liquid2 Ventures, FundersClub, and Jenny 8 Lee participating.

The new roster of investors join a list of prior backers that includes Y Combinator, 500 Startups, Christina Cacioppo, and Jack Groetzinger.

So what took them so long? Founder Ryan Bednar tells TechCrunch they wanted to take their time with the seed, rather than raise more money than they needed — a position that was possible thanks to already being profitable at YC Demo Day.

“I admit that this is unusual,” he says of the slow seed, though he also says they did raise a “small amount” after demo day, before filling out the rest this month.

“I saw many YC batchmates raising massive rounds pre product-market-fit, which can end up being a mistake,” he adds. “We probably could have raised a few million at Demo Day but ultimately didn’t feel we were ready for it. I didn’t know what I would spend the money on, and we were growing without it, so we chose not to. I wanted to raise capital when I felt we were ready to use it for growth, and now’s that time.”

Bednar also says he is “selective” when it comes to investors — and “specifically” wanted to work with Initialized, saying he’s “known Garry and Alexis personally for years, and trust that they would support us in building a long-term scalable business”.

Commenting on the funding to TechCrunch, Initialized Capital’s Alexis Ohanian tells us: “Even though so many businesses depend on traffic from search, it’s a challenge for them to be data-driven about SEO. RankScience makes it easy to test changes to your website that can lift search traffic. They also automate a growing number of technical SEO tasks, which otherwise would take engineers away from building product and infrastructure, which is really exciting.”

RankScience plans to use the fresh funding to hire more AI and machine learning engineers, with headcount growth targeted at its SF office.

While the founders have stepped back from pronouncing ‘the death of the SEO expert’, they are still touting the power of automation AI for SEO — noting how, after crawling a customer’s site/s, the software automatically proposes “SEO enhancements and experiments” to customers — for “one-click [human] approval”.

It also includes what Bendar bills as a “self-driving car mode” — where the tech will deploy the touted “enhancements and experiments” without customer approval. But he concedes it’s not for all RankScience users.

“For about half of our customers, we’re their only SEO vendor so we automate SEO services 100% for them, and for the other half, our software augments human SEO ability, either from in-house marketers or agencies,” he says, explaining how the team has evolved their thinking on automation vs human agency and expertise.

“When we launched we didn’t think hard enough about what sorts of controls SEO managers at larger websites would want, and we tried to automate everything without giving marketers enough control. This was a mistake and we’ve worked hard on correcting it.

“This should have been obvious but it turns out that SEO managers are highly selective about what sorts of HTML changes our software might make to their webpages. So we’ve spent the past year building tools to give SEO marketers complete control over everything our software does, and also advanced editors and tools so they can create their own SEO enhancements and run SEO split tests through the platform.”

For those who make use of RankScience’s ‘Self-Driving Car Mode’ the software is replacing SEO staff “completely”, but he adds: “This works especially well for startups and medium size businesses. But SEO is such a multifaceted problem, we want to give larger companies with marketing teams complete control over our platform, and so we work with both types of customers.”

As well as (finally) closing out its seed round now, RankScience is also launching a new self-service platform for startups and SMEs — touting greater controls.

On the customer front, Bednar says they have “hundreds” of sites on the platform now — and are serving “hundreds of millions of page views per month”. Cumulatively he says they’ve deployed “millions” of SEO split tests at this point.

“Our customers run the gamut from startups just getting started with SEO to publicly-traded companies,” he continues. “Our best industries are SaaS, ecommerce, marketplaces, healthcare, publishing, and location-based sites.

“We’ve recently been working with more consumer goods brands, and we’ve also launched a partnership program so that we can work with SEO and Digital Marketing Agencies and independent consultants.”

He says the vast majority of RankScience users are based in the US at this stage but adds that Europe is a “growing market”.

In terms of competition, Bednar name-checks the likes of Moz, Conductor (acquired this year by WeWork), BloomReach and BrightEdge — so it is swimming in a pool with some very big fish.

“Most of these products are more akin to advanced SEO analytics suites, and we differ in that RankScience is 100% focused on data-driven SEO automation,” he says, fleshing out the differences and RankScience’s edge, as he sees it. “Our software doesn’t just tell you what changes to make to your site to increase search traffic, it actually makes the changes for you. (Now with more controls!)”

Building a long tail of evergreen content

The first law of content marketing is that publishing is an expensive business. No matter how hard you try to circumvent it, the first law will always come back and bite you.

One common approach is to outsource content production, which can be cheaper in pure monetary terms. But unless someone, preferably an editor or with some editorial skills, is closely managing that process you’ll end up paying more for it in the long run.

That’s because the second law of content marketing is that crappy content generates crappy business. If the articles you’re publishing are just plain bad, then you won’t generate traffic. If they are good but only tangentially related to your business, then you’re going to be attracting readers who will never become your customers.

Just because you can publish to the world for free, doesn’t mean content
is free

Lots of startups approach content production as something that can be done in the founders’ “spare time” – at weekends or evenings. That might work for a while but it’s simply not a sustainable, scalable or sane strategy. And it requires massive discipline, because as a founder or early stage employee, time spent writing blog posts, hosting podcasts, or shooting videos is time you’re not doing other work that often seems more urgent or impactful.

Just because you can publish to the world for free with a Medium/Twitter/LinkedIn account doesn’t mean content is free. Content marketing should be approached and resourced in the same way that any other marketing function is, which means you have to invest in it if you want to break through the noise.

So how do you make sure you’re getting a return on your investment? Not in terms of how you measure success – the KPIs you need to measure will be unique to the circumstances of your business. But how can you ensure you’re getting the best return out of that expensively produced content?

Enter evergreen content

The strategy that has paid off – and continues to pay off at Intercom – is publishing evergreen content.

What is evergreen content? Put simply, it is content that addresses timeless themes and answers recurring questions people in your industry are going to have. Rather than reacting to recent events, it takes a broader perspective on topics. It is content that focuses on enduring ideas and advice that will have currency for weeks, months and years – evergreen content retains its relevance because it addresses universal themes.

When I worked as a journalist, evergreen had negative connotations. Real journalists were out breaking news, not reheating old themes at their desk. (Although that was before the internet ate their advertising revenues – but that’s a post for another blog and another day).

The content itself is highly temporal and has no value once the event has passed

Hardly surprising when most news organizations’ idea of evergreen content is “Everything you need to know about X event happening in our city” or “How to watch Y sports game that the whole country is talking about”.

The mistake is in mistaking an evergreen tactic for evergreen content – those articles will generate a nice bump in traffic in the day or two before X event as people figure out their travel plans, or when they are trapped at work instead of watching Y game, but the next day they’re trash littering your Google index.

At Intercom, the tactical equivalent would be to publish articles on how to survive this year’s Dreamforce or speculation on the next Facebook F8 announcement. The tactic might be age-old and recyclable (the preview, the round-up, etc), but the content itself is highly temporal and has no value once the event has passed.

A far more efficient and productive use of resources is to publish evergreen content that touches timeless topics and answers those recurring questions. (Sidenote: the rise of voice interfaces means if you want to improve your SEO you need to start thinking about answering questions not simply targeting keywords.)

Here’s just three examples of articles we’ve published that answer perennial questions:

Finding value in the long tail

Evergreen posts like these continue to generate traffic months and years after they’ve been published. In contrast, news or content tied to current events drops to zero the next day or whenever the event is over. Assuming both articles require a similar amount of work to create, it becomes pretty clear which is the more profitable strategy.

Not every article we publish here is evergreen but it’s certainly the bulk of our output. So much so that we actively monitor how much of our traffic is being generated by recently published articles versus evergreen content (defined as posts published more than 9 months ago).

The chart below shows how that panned out over the last 3 quarters – the bars are total traffic while the line is the contribution of evergreen content. This long tail effect means that 40-60 percent of our pageviews each month are generated by articles we produced at least 9 months ago.

Intercom Evergreen content
Intercom blog traffic over the past three quarters, with content older than 9 months indicated by the “line”

Tips for your evergreen content strategy

We’ve heard the argument that only well-resourced companies like Intercom can afford to follow this strategy. But actually it’s one that we’ve pursued since Intercom was a 4 person startup and was an early marketing success before we even had a marketing team. Don’t take my word for it: Tomasz Tunguz has done a compelling analysis of how an evergreen strategy can 3x your audience over a year.

If you think an evergreen strategy might be for you, here are our tips for implementing one:

  1. Don’t publish and forget Not to stretch the analogy too far, but just as a garden needs to be weeded and the dead foliage removed, evergreen content can’t be allowed get out of date. This is particularly an issue if you’re writing technology or business, where there’s a pretty relentless pace of change. You’ll need to ensure you have a program in place so that your older articles are periodically being reviewed and optimized.
  2. Hold your nerve You only get the compounding benefits of an evergreen strategy if you stick at it. Programs that run out of steam after a couple of quarters won’t reap the benefits.
  3. Shine a spotlight on it Evergreen is not just an SEO play. You should take every opportunity to get eyeballs on your best performing articles. New subscribers to our blog get a welcome email that includes classics from the vault, while our weekly newsletter includes an “In Case You Missed It” section where we highlight one of our previously published posts.
  4. Social is evergreen too You should also be comfortable sharing your evergreen content across your social networks. Some of our best performing social posts are when we share links to our older articles.
  5. Lean into it When we realised the power of evergreen we made the decision to double down on it. One small but telling example – we don’t put dates on our articles to make them more likely to be consumed long after their original publication date (although that puts additional pressure on us to ensure we are implementing tip #1)
  6. Convert visitors with CTAs Content isn’t free to create, so don’t give it away for free either! Make it clear what you think the next step for your readers should be, which will also nudge them down the funnel towards becoming a customer.

Evergreen is not the solution for everyone. But if your resources are limited (and whose aren’t?) it’s a powerful place to start.


Want to learn more about how to build your brand and sell more products in a non-spammy way? Download your copy of our book, Intercom on Marketing:

The post Building a long tail of evergreen content appeared first on Inside Intercom.

Data-Driven vs. Data-Informed

Many product organizations have stopped focusing on being data-driven, and are pivoting towards being data-informed instead. What’s the difference between the two? Let me illustrate with a real scenario. Product Study: Data-Driven Demand Assessment One of my past initiatives was to find product-market fit for a particular product idea that we had. The goal was to use landing pages, SEO (search ... Read More

The post Data-Driven vs. Data-Informed appeared first on Product Manager HQ.

Google’s updated Search Console gives you more detailed search performance and indexing data

 Google is launching an update to its search console for webmasters today that, among other things, brings redesigned search performance and index coverage data to the service. In addition, the console now also features more detailed data about the errors that a site’s AMP pages and job listings may be throwing (and how to fix them). The company had been testing these updates in beta over… Read More