I recently got some feedback from one of my top alumni on a presentation I gave, she said “Wil, I’m following you on this “We’re going to win for Humans” thing 100%, but what do I do to prepare my team tactically in the next 6-12 months?”
I took that feedback, reflected and thought…you know what she had a point.
My newest "ranking factor" has been can I get humans to care first. I talked about this here in this podcast.
I laid out a vision for where search + AI is going, but didn’t provide what the next 6 months could look like.
Your clients, managers, and CEOs are stuck trying to hit goals now.
That puts you in the position of balancing today’s KPIs with tomorrow’s activities.
Your customers are not using AI to answer questions en masse, so we need to see the future, prepare for the future and still live into today’s reality…people are Googl’ing still. It’s a balancing act.
Here is what I would be doing for the next 6 months, some involve AI, some don’t. Earn trust by admitting that customers are all over the place, attribution sucks, and the SERP is crowded more than ever, reducing the value of SEO.
The ROI of $1 invested in SEO is not what it was 5 or 10 years ago, you need to be honest about that or lose trust in the c-suite, there are many ways that you should account for all the value you could be generating, which that post is a dang good start.
I fundamentally believe that more than ever in my 20+ year history of marketing on the web that this is the time that your ability to win with humans (not their proxies, cough Google, ChatGPT) gets the greatest reward.
Be honest:
SEO wins ≠ Human wins
I recently Googled Whatsapp Statistics 2024, saw a site I trusted, its title tag said 2024, its publish date said Dec 2023, all the SEO signals pointed to “I’m about to get a good answer” then I saw this chart as the first data image, data is 4 years old. This is the friction I showed earlier, this hurt the brand in my opinion, this is all the reason why search is broken, and LLMs don’t do much better in this example either.
#1 - get your credit for human wins
You might be winning over humans with your content to pursue content that wins over Google, which is just a proxy for human trust anyway. This set of minor tactics to look good for Google, could be leading you down the sea of sameness issue, causing you to never stand out.
Today you should be looking for content your company has built over the last 3+ years that earned links, social shares, private shares in places like Whatsapp, Slack Communities, Teams, etc.
This content is showing you that yes, you have written content that might be “winning for humans” where you may be getting no credit for the leads, signups, etc you are getting, as those KPIs are not rankings.
Here you can see our LinkedIn traffic is growing after a pretty serious change in our approach to the platform.
Most SEO professionals would not get credit for all this traffic coming from blog posts that are taking off on social, I wrote a good bit about how I see social as a key to content success in the future.
How are you accounting for influential people linking to your product, via LinkedIn? What about engaging in the comments and linking out in a way that helps people better understand your product?
#2 - Track Google Layout Changes
Track Google layout changes in excruciating detail using a tool like Visual Ping to help you keep track of Google search result changes - Here is how easy it is to track layout changes. It is CHEAP, for about $10/mo you can check 1000 pages, and a page could be a SERP. Don’t wait for your boss to give you budget, you need to do this today to defend your value, today..
Google layouts right now are in a constant state of flux, and that is going to cause your click-through rates to move a lot, that is going to affect your throughput for the rankings you’ve worked so hard to get.
That leads me to the next thing you should be doing in the next 6-12 months.
#3 - Revisit your CTR projections quarterly
1000 MSV with you at #1 could be 10% CTR today driving 100 clicks, lets say you get 2 conversions for every 100 clicks. In a world where your CTR drops by 50%, you get 1 conversion, how does that impact your projections for the year?
If I was paying my SEO firm $6,000 per month, and I used to expect 6 conversions from your rankings. I used to get a trackable lead at $1,000 per lead, that is now $2,000 per lead - is that sustainable or worth it?
If the Google layout changes, and we lose 30% of our clicks on rankings 1-5, how off would we be from our current projection on sales / leads?
To me you should be able to answer this question “in the hall” passing by your CMO.
I like VisualPing for sharing that with executives because it's very easy for them to see (thus Visual, ping). For an enterprise SEO, you would probably want to use a tool like Nozzle or Data for SEO that will give you where SERP features show up in terms of how many pixels down it is on the search result page.
To me, the most important thing you can do right now is to understand how the layouts are affecting click-through rates, which will impact traffic, which will impact goal attainment.
This is happening NOW, AI-search isn’t, so focus here so you are never caught flat footed.
#4 - Find keywords that are resilient to layout changes
Broad based CTR impact curves don’t work for us. CTR curves by n-gram or intent is a more advanced search marketer IMHO.
If “What is” or “How to” keyword CTRs drop 50% with AI overviews, fine… which ngrams or intent types don’t drop when AI overviews (or any other SERP feature) show up?
In other words, when humans add certain words to their queries resiliency is impacted positively or negatively I would want to know that. You might find that moving from #4 to #2 is more worthwhile on queries where a new SERP feature is present since the keyword is more resilient.
What CMO’s and Executives will want
If you want to level way up, have a plan that shows for every 10% drop in traffic for the same rankings, how you will use search data to create more value in your org, or how much faster the org needs to operate to get things done.
Imagine a world where you used PAA’s (SEO data) to improve paid ads, social ads, and email headlines?
We’ve been talking about this for years at Seer, but to me if the ROI of SEO “drops” then I’m still relevant because I can try to backfill that direct value, with indirect value in supporting other channels that are growing using search clues.
If humans are getting more answers from Linkedin Pulse and Reddit - then go learn how to assess where there’s risk in your business.
We recently did an analysis that shows when paid keywords (which gets us conversions) have Reddit showing up in the top 3 spots. We were then able to plot out Reddit visibility (any SEO can do this) with Reddit visibility on converting keywords.
#5 - Keep optimizing Google
Seems obvious, but it needs to be said, most of your customers are here. I dunno if your best ones are but MSV shows people are still mostly here. The big news is the Google ranking factors leak or whatever we want to call it. I think for most people in SEO, there could be enough there that in the next 6 months, you might be able to pull out something that could alter your SEO strategy or at least to test.
Well, that isn't what my area focuses on, I could see there being some value gained if you've been needing to validate hypotheses that you haven't been able to get buy-in on.
#6 - Get better at convincing folks to execute!
The skill of convincing your leaders to invest in helping you execute on hypotheses will be valuable in any marketing avenue. For SEO / Content folks, if you are losing CTRs, then you need to improve your execution game, you might be able to overcome the drops in CTR by getting more done faster, but you gotta make sure the stuff you are prioritizing is worth doing. Find out what language people in your organization use to get budget or resources to execute faster than you are.
Book recco: Persuasion & Selling the invisible.
#7 - More conversions with less traffic - study CRO.
Make better friends with your CRO pals, or heck start studying CRO now. Which of the following 2 people would you want to hire when your traffic from SEO is down 20%?
Now, let's talk about AI, specifically the AI overviews.
I think it's really important for you in the next 6 months to try to understand where Google is going to show AI overviews.
Sure Google has pulled a bunch of the AI overviews back as you can see in this report. Looks like as of today (6/25/2024) they are starting to come back in healthcare according to Lily Ray’s data.
#8 - Google’s AI Overviews pullback is a GIFT!
But I think that is only for now, as they will review, reload, and re-launch these features in the future.
This Google pullback is an opportunity, you can now see the impact of these overviews on organic, paid, brand, etc, as Google just gave us a moment in time where we can see our metrics during this phase so we can help prepare our orgs for what this might look like in the future.
Ziptie seems to be one of the first companies out with a product that seems pretty robust for enterprise companies where you can track what keywords AI overviews are showing up for.
Understanding what kind of search terms and what kind of parts of search terms trigger AI overviews will be a very helpful component of your strategy because again, it helps you to model out potential impacts in the future 6 months or year for your company so that you can help them to make better decisions.
#9 - Risk dashboards for Google AI Overviews
Aleyda Solis built this lovely risk dashboard for you already...
In our data product Seer Signals I built a visual that looked at “How to” and “What” keywords across every client's paid search terms for last month.
That allows me to see what % of their conversions come on how to or what (or any keyword we hypothesize), then if we find that AI overviews show up for these keywords more, we know which clients are most likely to be impacted with their own data vs some benchmark.
As you can see here for one client 34% of all of their conversions came on “How-To” keywords. Their impact from AI overviews might hurt them a lot more than another client.
They deserve a different strategy than a client who is at 3%, that is why generic studies don’t work for us at Seer and why we built signals for clients and in house teams.
#10 - “I” queries in your SEM search query reports
Build a graph that shows % growth over time in queries starting with I, from your paid data.
Don't have search terms due to the PMAX tradeoffs? - Then Mike Rhodes has your back - get them search terms back with his script, accept no excuses.
Go over to ChatGPT how many of your queries start with the word “I”? I know for me a TON of my inquiries start with “I would like to know…” cause that is how we as humans actually think. Search engines have taught us to not use the natural language in our heads.
Anyway…in Seer Signals we’re tracking the growth in queries starting with the word ”I” for every client, every month. If you have access to paid data queries, pull this and trend it out, the growth could impact your need to increase your investment in AI search / tracking / etc.
Warning: Don’t think about optimizing for AI results.
I think trying to figure out how to optimize for searches on generative AI is not something that the average SEO should do. I would recommend you focus on how AI overviews showing up are affecting things like my click-through rates for the next 6 months.
Optimizing for these is a lot like the SEO leak (although I would much rather invest my time here), for now the ROI for you testing isn’t worth it, but you should be monitoring heavily.
Britney Muller is my GO TO on helping me recognize my own ignorance in how these things work, looking forward to more content for her to share from our private chats.
You can do this for the future, but for the next 6-12 months, I wouldn't touch it.
#11 - Compare PAA's vs AI answers
We’ve talked a lot about how to track brand answers in LLMs using Poe.com and in ChatGPT for sheets to catch hallucinations and what these LLMs know about your brand.
You need to build the proactive dashboards that will alert you to when your customers could be moving over, so you can better time when your skills should shift. You need analytics data (referrals) and visibility data (PAAs vs AI Answers)
One of the things that I would do is I would take questions that Google shows in the form of "People Also Ask" and run those and other questions through multiple models, determining what our visibility as a company looks like for similar types of questions as "People Also Ask."
I would compare my visibility today in the search results and "People Also Ask" questions against what my visibility looks like if somebody asks those questions to an LLM.
We have this in Seer Signals in Alpha now. This lets us look at "People Also Ask" questions and which companies show up at scale.
I’m fully aware of how much the answers change, but I’m not going to be caught flat footed when I get asked tough questions from a CMO.
C-level's don’t want to hear “the answers change all the time” they want you to have a POV, and here's what it should sound like...
I’d much rather show up saying “We ran this question through these 4 LLMs, 100x each and this competitor shows up 50% more than you do”. Then say… “this changes all the time and we gotta know that this could change in a month, or that adding the word enterprise to this query can completely change the answer.”
#12 - Take one task and cut time time by 25% using AI
The other side I would be looking at is automating or speeding up part of my job using AI. At this point, I think it's pretty safe to say that AI is here to stay to some capacity in how we do our work, and I would want to become a go-to in my organization on that front, check where you are by looking at this AI survival kit for marketers. I would want my managers to see that I have done dispassionate reviews of how my work is being done with and without AI and where it creates value.
Your leaders are as worried about being seen as behind in AI as you are, take that as an opportunity to help them, become their AI whisperer, where they feel comfortable sharing struggles.
#13 - Reddit analysis and action
Remember people aren’t suddenly less inquisitive, they are just not going to the same channels to get their questions answered. The big spike in Reddit traffic is an opportunity for you to better understand humans, the questions they ask, and how people answer them. For now, start analyzing the impact on conversions, but be ready to offer a POV on questions like:
"Should we join in" or "should we advertise there".
If you think this isn't SEO, you are right...but it is showing up where customers are, for now and I recommend you get used to being flexible in your title, and follow the customer instead.
You should be looking for how much Reddit has increased for your keywords, our preference is to use combined PPC and SEO data to determine if Reddit is showing for converting keywords because well, you need conversions 🙂
#14 - Improve your interviewing skills (18 months)
I’m starting to get my spidey senses up that humans are going to want more answers from humans than algorithms. Whether that is search or ChatGPT, doesn't matter. Gen AI removes the site after site friction, but I dunno if the answers are good consistently enough.
Given that, I think we’re a long way away from Organic Search AND Organic Social being a single job post, but I do think at some point your best organic marketers will also be able to distribute great content through social accounts and would know how to work with human SMEs to promote.
Cause we all know by now… Organic visibility on LinkedIn accounts is DEAD, SMEs (influencers, yuck) will rule, go find them, start to know what makes them tick, become one of them, but people want deep answers from other people.
I would start leaning into ways to pull insights out of my subject matter experts in my company. You might want to start playing with Vapi.ai - I got a video coming on this.
You may ask, "Why would I want to spend my time in the next 12-18 months asking subject matter experts in my company about their thoughts?"
To be a damn marketer. That is why. You can't just put a bunch of keywords in anymore and rank and hit goals, you might have to make friends with other people in the org who know things, the horror.
SME's are your unlock to breaking out of the sea of sameness.
The reason why is I believe that more and more people are going to end up getting basic answers from LLMs, but more advanced answers from thought leaders, I want to be in the business of extracting from thought leaders, their ideas, so I can help them become…influencers. Eh, more influential. That is where I would start making my longer term investment, because like it or not, we’re all influencers now…the question is how influential will you be as your customer moves where they get their answers from?
