beyond google algorithm changes

Google has just changed how they treat redirects.

That doesn’t sound all that exciting, but in the SEO community, it’s kind of a big deal.

It used to be that 302 redirects – that’s the temporary kind – did not pass any link equity, only 301 permanent redirects did that. No “juice” flowed to the new page when the wrong redirect was used, much to the chagrin of non-technical web site owners.

Now, with Google’s change, that’s less likely to happen.

Google has been making those types of clean-ups pretty regularly for the past decade.

SEOs are still crucial in today’s world, but for a slew of different reasons. The old skills traditional SEOs possessed just aren’t that important anymore:

  • Inserting meta keywords to optimize a page used to be a thing. Google found the field too spammy after a while, and stopped looking at the field altogether.
  • Inserting the target keyword into the content multiple times used to be a thing. Now, Google punishes sites with low readability due to marketers spamming keywords on the content.
  • There was a point when exact match domains had a high impact on search results, leading to some SEOs registering multiple non-brandable domains and using those for their outsized impact on the results page. That impact is smaller now, and Google is heading in a completely different direction.
  • Getting content in banners read by spiders used to be a priority, with Flash and javascript mostly being invisible to search engines. Now, Flash is gone, and standard javascript banners readable to spiders, SEOs rarely need to step in to fix that from a technical standpoint.

There’s still a place in a ton of organizations for “technical” SEOs.

There’s getting tags right, so the results on Google’s pages are richer and more people will click through. There’s still making sure the site architecture is right and the sitemaps are in place for the pages to be spidered.

But the purely technical aspect of SEO used to be bigger. There was a point, some time in the recent past, when technical SEO seemed to be the only game in town – that’s very clearly not the world we live in today.

Today, you need web usability knowledge, conversion rate optimization, and a range of other skills – they are table stakes.

A Changing Landscape

Some marketers are asking exactly how that happened.

What happened was a combination of two things:

  • Marketers targeting where Google is rather than where Google wants to go
  • Technology improvements at Google

All of the things that stopped being as important as they were are inputs. Meta keywords is an input – it’s about telling machines what search terms are important to a page. A 301 redirect is an input – it tells the system what kind of transfer should be done, whether it’s permanent or temporary.

Markups for banners are inputs. Meta descriptions are inputs. Browser page titles are inputs.

They are about getting machines to understand the content.

This was important in the past, when Google spiders needed all the help they could get.

Google’s 2016 spiders are much, much, more sophisticated than those from a decade ago. They require less help from web site owners to do their thing. So something else is becoming more important – the output. If the input is the ingredient, the output is the dish, and how happy users are with the dish.

Here’s what great output looks like:

  • Visitors don’t immediately leave a site after landing on a page
  • A relatively high percentage of visitors convert
  • Web users are happy with the overall experience

Today’s world is as much about the inputs as the output – and the future, well, with deep learning picking up steam at Google, the future is going to be lighter on the inputs, and heavier on the output.

What do deep learning and machine learning mean for marketers?

Click here to read Google, Machine Learning, and Marketers: The Least You Need to Know.

An Output-Driven Approach

It isn’t just that the spiders getting smarter. The actual algorithm is getting ridiculously smart, too.

Google is diving further and further into deep learning, which means that in the future, the machines will be observing what works to make users happy, and using that to rank pages.

Today, that’s not really the case.

Engineers from Google’s team can take a factor like mobile speed performance or root domain links and tweak those to make them more or less important, then check if users like the pages that rank.

In the future, machines can take hundreds of different combinations of factors, and then use some combination that works to make the users happy with Google and performing more searches. That combination doesn’t necessarily have to be intuitive to marketers – the sites that rank can have a low number of overall links, low shares, no tags, and still rank.

At that point, any individual engineering team within Google may not know which factors are weighted heavily, or if some factors are used to rank at all. W what they will know is the number of users who search, and who don’t go back to the search page after getting to a web site. What the engineers will know about is the outputs, and how well the machines are optimizing for that.

Go Where Google Is Headed



It has been very profitable in the past to look for flaws in Google’s algorithm.

Marketers have targeted how mature Google has been, and it has paid off. Meta keywords mattered, until they didn’t. Exact match domains ranked, until they didn’t.

Lately, though, with Panda, and Penguin, and Hummingbird, it’s been tough going for marketers who are valued for their ability to look for cracks in Google’s algorithm. Those cracks have been showing up less and less, and lately, Google has been pretty aggressive about weeding out marketers whose specialization is Google-gaming.

If you care about your rankings in the future, you should invest in skills that target where Google is headed:

  • Sites that have good user experience and can make people stay (skill set: usability testing, customer interviews)
  • Pages that convert well (skill set: split testing, conversion rate optimization)
  • Pages that give users what they need, quickly (skill set: exit survey number-crunching, page speed optimization)
  • Repeat visitors, trusted sites (skill set: general web usability, marketing automation, client nurturing)

To be clear, it still pays to know how you can have reviews and other types rich information show up on the search page itself.

You still need to know your way around things like HREFLang, canonical tags, and other SEO-centric information. But if your differentiator as a marketer is that you’re familiar with those elements, it’s time to expand your skill set.

Google is heading towards a future that’s heavy on the output.