There are many types of SEO practices, from technical, to content and link-building. How do you know which one’s best for your online business? If you really want to see a good return on your investment, Wil Reynolds believes you have to start from a place of data.
Wil Reynolds is the Founder and Strategy Director of Seer Interactive, a company that he set up on his own in 2002, offering SEO, PPC and Analytics services for all types of verticals, from pharmaceuticals, to e-commerce and SaaS. Wil has 15 years of experience in Search Trends and is regularly invited at the most important Marketing Conferences around the world to talk about the Future of Search and Content Marketing.
Wil came to keynote for the first time in Romania at GPeC SUMMIT on big data and e-commerce search marketing. In the interview below we talk about the main ideas he delivered on stage.
Wil Reynolds: I think the number one thing that I was trying to convey to people at GPeC SUMMIT was that, a good chunk of your spending paid and on those ads, the paid clicks, the shopping ads, are happening on really small micro-search keywords. And if you ignore them, very often they’re the ones that are getting a lot of your conversions, and usually they’re costing you 30-40 % of your spend.
But yet, we kind of are allowing Google’s AI to automatically manage those words. And today I wanted to show people some missteps in how Google manages that, and to make sure that people are aware of how to check up to make sure that Google’s AI and automation is actually working for them.
The best example is the turkey example I showed today. Google has different ways of saying “We’ll match a closed phrase to that phrase, because we know that it’s what you wanted”. Generally, I think that’s a good thing. However, if you’re a client that sells turkey meat, you probably don’t want to be showing up for words that are Turkish, it’s a totally different thing. And it seems like Google’s machine learning missed that one for one of our clients. And our client is getting clicks on those ads where they’re saying “We sell turkey”, while people are looking for Turkish clothes, Turkish coffee and clicking on my clients’ results, costing them money.
GPeC: You previously said that ranking is not enough to drive success. Suppose you were the owner of an online shop, what would you focus on in your marketing strategy, considering that SEO is not enough?
Wil Reynolds: It’s funny, we had this belief that, oh, if I rank, people are gonna click. And I think the thing that we forget about, especially with brands in e-commerce, is people have trusted brands that they know already. In the US it might be Walmart, Home Depot, Amazon. So, if people are looking for those brands and see your company up there, I’m almost like “Well, I’ve gone to Home Depot before. I know how far they are from my house. I know about their return policy”. So, even if you outrank Home Depot, in my opinion, for certain products, there’s a good chance that I’m not gonna click on your result anyway.
So, I think it’s really important to build up on your brand, so that you start to become trusted or, otherwise, I might just skip over the top result that might be you, to click on somebody who I know has a good return policy and a really broad product selection, for instance.
GPeC: In your opinion, what would be a top three link building approaches and technical approaches?
Wil Reynolds: I think that technical SEOs are going to solve most problems with technical SEO solutions. And link-building SEOs are going to solve those same problems with link building solutions. And content SEOs are gonna solve those problems with content. And I think what we’re missing is that strategic layer and the data to tell us when we might not need to use a technical implementation, because people are ranking already without having it.
But a technical SEO is gonna be like “You have to do these things!”, or a link builder is gonna be “We need links!” And you’re like, “Well, how many do I need?” Well, I don’t know how many you need to rank for this word. As many as possible”. And then you’re like, “Well, these other sites are ranking above me without any links, so how? Help me to understand the value”.
I think that we too often lead in with technical, content, etc. and we need to really start from a place of data and say where does the data show me I need links. Where does the data show me that I need to fix this thing technically or that I need content? Because right now, I think, myself included, everybody has their own leaning. So I’m more of a Content-style SEO, so I’m gonna run in and solve most problems with content. But then, I didn’t pull the data to see if maybe I should have started off solving it with links or with technical.
So I feel like we have to start from a place of data to determine what type of SEO we need, or if you ask a technical SEO, you’re gonna get a technical SEO solution. You ask a content SEO, you’re gonna get a content solution. If it’s a link building SEO, you’re gonna get a link building solution, and none of them are looking at the data across other divisions to say “Maybe you do need technical and not me right now”.
GPeC: But what if we are speaking about an online shop that’s just starting in this business. They don’t have any data in the beginning.
Wil Reynolds: What you should do is start off paying for the data. I say start off paying for the clicks, because there’s so much value in the clicks that you get from Google, so then I can see, “OK, I paid for these words, how many people clicked on three or four pages? What search terms led people to adding to cart?”. Even if they didn’t convert, you can start to say “Wow, these terms led to a lot of people leaving”. So now I know: don’t build content for those words yet, because all those people bounced.
But you have data now to tell you not to go after a word, versus “I think that microphones are important”. Shouldn’t we go after the word microphones for three or four weeks and see if those people actually elicit behavior on the website that says they’re interested in buying, before we go and spend months and months of resources to build out new content and links and do technical? And most people are like, yeah I probably should spend the money for a couple weeks to get the data to tell me whether or not I should do that.
GPeC: Studies predict that by 2020 50% of online search will be voice search. From your experience so far, what are the steps a business needs to take in order to optimize for voice search?
Wil Reynolds: I don’t know how much voice is or isn’t going to hit those numbers. Right now there’s no way. We are six months away from 2020. Do you do half of your searches today via voice? Most people aren’t there yet.
What I tell my clients to do is, let’s get all of your branded search words for your brand – so these are people who know your brand – where an answer box shows up on Google, and see how often are we not the answer. Because that’s really bad if, when somebody says, “How late is your bank open on Fridays?” and the answer is coming from another website, that’s really bad!
So my first step is let’s use your paid data, look at all the different searches people have done with your brand, see which ones are triggering answer boxes, and the ones that are triggering the answer boxes where you’re not the answer, fix those answer boxes first, because they’re going to be mismatches for voice.
GPeC 2019 is an event powered by FAN Courier
Gold Partner: VTEX
Brought to you by: Orange
Official radio station: Kiss FM
Official hosting provider: Hosterion
Limousine Service: City Limo