For episode 154 of Search Engine Nerds, I had the opportunity to interview Keith Goode, Sr. SEO Strategist at IBM, on how data plays into SEO and digital marketing.
Goode also highlights the importance of looking at data to “get the answers that we really need – not the answers that we want” and shares insights on democratizing data across all disciplines.
What is your purpose for being data-focused at IBM right now?
Keith Goode (KG): I don’t really know how you can be an online marketer and not be focused on data. The problem I find with a lot of folks that work with data is they want the data to tell a story that they want to tell.
They don’t stop and ask the data what’s the story it’s trying to tell. We try to, “OK. We got some bad stats here on analytics. How can we spin this into a positive story?”
I’ve actually done this myself in the past. I’m a little embarrassed to admit that, but I think a lot of marketers do that.
They see data and they’re like, “Oh man. This doesn’t look as good as I’d for it to look. I said we’d go up in ranking and we’ve gone down in ranking. How do I spin this data to really tell a better story on my behalf rather than actually saying what the actual is?”
We’ve really embarked at IBM especially… where we’re really trying to democratize data across all the disciplines and as much data as they need to gain access to.
The reason for that is we want to be able to eke out all of those individual stories that the data is trying to tell. Maybe even look at data points that we wouldn’t normally look at, and get the answers that we really need – not the answers that we want.
The decisions through data is a program that we have at IBM. This is tapping marketers into as much of the data points as they can potentially get their hands on.
This is across not only just organic. We’re interested in getting all channels involved and taking a look at all of the data.
We know from a historical standpoint that, synergizing your SEO and your PPC efforts really is a “one plus one equals three” type of benefit.
We definitely don’t want to limit our paid folks to just whatever our paid analytics platform is telling us. We want them to dig into the organic data as well.
We want them to dig into the user stories and the user metrics, how people are interacting with the site, verbatim feedback on the site. Those types of things can help us across all the channels, not just organic.
Brent Custoras (BC): Over the years, we’ve actually seen a little bit of that data taken away from us, through privacy and [other factors]. Regardless of whether it’s right or wrong, that data is starting to become harder to interpret.
With that, we’ve found that not only have we lost data, but we really haven’t filled that void… I’m curious, are we evolving with data and data collection or we’re kind of in the standstill?
KG: It’s an annoyance for me as well when Google was, at 2013, took away our access to the keyword data and analytics. Here we are six years later…
Of course, there are reasonings for hiding that data were, quite frankly, quite a BS… It’s not a privacy issue. It’s data.
Data is the new natural resource. They want to find ways that they can sell it. That does irritate me, but you can’t get stuck in that quagmire.
You can’t say, “Well, I don’t have access to keyword data, so I can’t really tell you how well we’re performing, because here’s what we know.”
How do you get your keyword data?
KG: We do get it from our paid [Analytics]. We do pull it in from Google Search Console. We pull in as much as we can to start making some broader decisions.
What we’re really more interested in is not what keywords are driving traffic to us, but is the traffic that they’re driving, getting what they need?
We really move away from an inside-out type of marketing approach to an outside-in approach, which means that we are stewards of the Internet for the people that are trying to find our products and services.
We have to figure out what it is they need or we can drive all the traffic we want then. If it doesn’t convert, it’s not going to do us any good.
[At] IBM, [we’re] interested in showing our value than we are driving traffic… We moved away from focusing on keywords to focusing on topics.
- What are the bigger topics?
- How are people talking about those topics?
We can use keyword research for that. We can go out and we can take a look at a site like AnswerThePublic and see how people are asking about these things.
We can go to Quora. We can go to any number of these other channels and figure out how people are talking about the topics in general.
The keywords only give us an indication of how we should be also talking about that topic.
How should people determine how to set up data collection appropriately based on their goals?
KG: Setting up analytics for any organization, I think there need to be these conversations leading before you even get to that point.
If you’re talking about something along the lines of keywords, search volume is not going to be your only indicator. You have to decide, as a business, what we’re relevant for, what makes sense for us?
Here’s a really easy test to figure out if you’re right for any specific keyword.
Go to Google. Search it. Look at who’s currently ranking right now.
Is that even close to what you do? If it is, that’s your keyword. If it’s not, you need to get a little deeper go longer in the tail for that particular search.
A lot of these questions have happened before you even set up analytics. What I really encourage folks to do when the finances allow it is investing in a platform of some sort – whether it’s something as simple as Moz and their platform, or if it’s something that’s huge and complicated as BrightEdge, Searchmetrics, Conductor, or seoClarity.
I really highly recommend that when you’re setting up your analytics, you do it in a way that you can unify the views into a single view. You can pull in multiple data points to get to your conclusions for this.
You’re really getting into business intelligence at that point when you’re aligning across the multiple data collection sources. When you do that, you start to see a clearer picture.
You can use data initially to look at opportunity. Then measure yourself against those expectations once you set up analytics.
Then you have to look at the data even deeper and figure out what is the story it’s telling me is right now. What are other opportunities I may not have yet explored?
For example, we are mostly B2B business at IBM. What we have found in many cases is that it’s not a single visit that results in an engagement. That engagement might be someone submitting a lead form, making a purchase, any number of things.
That ramp up is about six months. There are about 15 touchpoints between the initial visit and the time they actually make a decision.
We have to have data that account for all 15 of those touches and figure out why did it not happen on the seventh touch point.
What happened there? Was that a bounce? Was that a miscommunication? Why did it take 15 touches?
Can you give a couple of examples of actionable data?
KG: Actionable data, for me, would be if I’m seeing a reduction in the number of impressions reported in Google Search Console but I’m seeing an increase in a number of clicks, I’m reaching the right people then.
The other thing would be that if I’m actually seeing a reduction in organic traffic, in general, but I’m seeing an improvement in engagement, that means I’m driving the right people as well.
[Y]ou can’t just stop at a single data point and say, “Well, we’re screwed. We’ve lost traffic. We’re no good.”
You have to look at that engagement metric. I got to be honest here. People say they want to drive more traffic to their site.
But what they really want is they want to drive more conversions to those sites because that’s how we pay our bills. We don’t pay our bills based on visits unless we’re running a CPM model of some sort.
Would you recommend somebody really get a data scientist?
KG: Yes, if you can. If it’s in the budget, definitely get yourself a data scientist on your team.
If it doesn’t make sense, just get more of your folks to look at the data on a regular basis. It doesn’t have to be just the organic team. In fact, I think you’re limiting your effectiveness by only sharing that data with the SEO folks.
Connect with Keith Goode on LinkedIn or visit his website to learn more about his work.
To listen to this Search Engine Nerds Podcast with Keith Goode:
Visit our Search Engine Nerds archive to listen to other Search Engine Nerds podcasts!
Featured Image: Paulo Bobita