The only link building you should be doing

The only link building you should be doing is internal. Internal link building (ILB) demonstrates an awareness of how content fits together across a site, across tasks for your users, the purchase consideration touch points, and, of course, search intent.

The exact opposite of ILB is external link building (ELB). By ELB, I do not mean “link earning” or natural link building, I like those things! I just that means those efforts extended to actively “build” (create) a given link from other sites.

The problem with ELB is that it “works” for getting traffic. But the more you focus on moving the needle of external links to a piece of content, the weirder and less relevant that content you have to create to facilitate your ELB gets. You start designing content to get links, instead of being focused on content as supporting your intended audience.

A prime example is this car title loan company’s “discovery center“, where you can learn about the “scariest urban legends by state” and “tailgate finger foods” all conveniently on one car title loan company website, resulting in what appears to be respectable search traffic numbers:

Filter on top 10 keywords for “loan” and “title” excluding branded terms, and we actually still see a lot of relevant first page rankings in Google. Hard to argue with, right?

The problem here is that these vanity numbers are completely misleading. This tactic is expensive, time consuming and demonstrates, compared to ILB, an incredibly low ROI that shrinks with every new piece of content released.

You can focus so much on “domain authority,” that you miss how wasteful this is for what really matters to your business. Google has no “domain authority” factor. It’s a made up metric by SEO tools as a proxy for a generalized PageRank algorithm that has long since been extricated from their org traffic algos.

So why do we still hear about DA? I’ve thought about this and believe its because:

• A simple score normalized to 0 to 100 is easy to understand for non-technical stakeholders
• It can be useful as a leading indicator / predictor of overall rankings of a site, or sorting top to bottom for a big list of potential outreach targets or partners
• It looks good in reporting as it increases over time, typically because it’s easy to manipulate and as your site gets bigger, it gets more spammy links, and DA type scores aren’t great at controlling for spam that Google ignores with ease
• Graph algorithms can be intimidating to understand and no one wants to look bad by questioning DA when they can’t defend the case for dropping it from their KPIs
• G is good at sowing fear, uncertainty and doubt and we’re forced to operate in a layer of obfuscation as we decide how to optimize sites
• Confusion lies in there now being something a topical PageRank-like component to how G evaluates links. So it seems like they still use PageRank. I know, it’s complicated. More here from Cyrus Shepard.

The point here is that external links to pages about siamese cats that you then link to your pages about power tools have minimum value to you. And those pages and their inbound links raise your link profile’s overall risk over time.

G could potentially further suppress blatant, easy to spot tactics like this, maybe tomorrow, maybe in years. Clearly a bunch of links to a page about which cities have the most billionaires should not score you authority points with G for title loans. And just the idea that a company would attempt it makes me, at least as a consumer, lose trust in that brand.

The “ELB works” defense does not hold up when compared to other ways that time and money could be invested. For every new piece of totally random but “easy to build external links to” content, the returns will diminish. Remember those impressive relevant first page rankings for the title loan company? Here they are over time:

Not so sexy now. The wayback shows they had about 45 of these assets in Oct 2017, just as their relevant traffic was peaking. Now they have 185. Their top 10 rankings for relevant terms is down 60% from 6,500 to 3,900.

Meanwhile non-relevant terms about the best hot dog toppings you never heard of or whatever the heck are doing really well and so stakeholders paying for these ELB campaigns probably don’t know why new leads are down when their overall traffic is up, or more likely, they don’t track new vs repeat leads well enough to know new leads are down.

The cost per piece for this type of link building was probably pretty high for them – in 2016 you could have paid $20k for just one piece of content plus link campaign. As tools and competition have commoditized this approach, market rate is prob down to something like $5k/piece average. That’s 140 pieces x $5k = $700k.

Do you not think that $700k could have been better spent creating authoritative treatments of relevant content to the type of people seeking to take an extra loan out on their car or motorcycle? Then all link building could have been relevant. Each new piece of content could be linked as it made actual sense for the user with old content. No link building budget or random infotainment creative required. And the moat they could have built between them and everyone else in terms of relevant traffic, branding, and leads would have been so severe those competitors would probably have to resort to Geico style commercials.

Which brings us to our main point: the only link building you should be doing is internal. When you don’t have a good piece of content to link to or from, that’s a strong signal you need to create that content. The gap analysis a strong content audit can provide will uncover those opportunities for you. Then your internal link building and content editorial calendar can be in lock step, and you get to build something that will only compound in returns on relevant traffic and conversions over time.