B2B brands should also worry about AI search
AEO is not just for B2C. Here’s how B2B marketers should think about the impact of GenAI on winning customers
Programming note: I’ll be off next week while I’m on vacation, so I’ll have another post up on August 29th.
In a post a few weeks ago, I talked about the potentially large impact of GenAI on search for B2C. Well, if you are a B2B marketer, you are not safe either.
It’s also important to note that ChatGPT is coming out with a search engine functionality called SearchGPT. You can read a review of it here. For now, this seems like Perplexity in that it returns search results in more of a narrative form, and you can chat with the results. But by the time it hits wide release, it may turn into something more.
I’ll also note that Gartner predicts1 that search traffic will fall by 25% in 2026 due to people migrating to AI search.
For B2C, the basic issue is that instead of people going on Google to find products, they will go on ChatGPT, Claude, Perplexity, etc. Those tools are very willing to recommend specific products, and nobody knows how they come up with their answers. Optimizing your brand for those systems is called AEO (Answer Engine Optimization) and is a new art that SEO (search engine optimization) agencies are trying to master.
For B2B, the issue is more complicated as I’ll explain below.
Direct sales is the obvious impact
The world of B2B sales is vast, ranging from a sole practitioner buying paperclips to a Fortune 100 company buying a $100M IT system or an airline spending $100M on a new plane or hiring a law firm. Obviously, the purchase process varies widely for these different types of sales.
To start, the obvious situation is B2B products that are sold through websites. This could be physical goods like office supplies or accounting software for small businesses. This is almost identical to the B2C situation. People will go on ChatGPT and say, “I’m starting a new business, and I need accounting software.” The response is pretty consistently
The situation is quite similar to B2C: to drive sales, brands will want to optimize their presence in these searches. Also, like B2C, brands will want to be positioned well in these results. If you work at Xero, you might not like seeing “Learning curve for new users; higher cost for premium plans.” Does that come from online reviews that the system has ingested? What do you do if you’ve fixed these issues and ChatGPT hasn’t updated its answer?
Perhaps in the future, the model will update more frequently and consider more recent reviews. That would allow firms to address these issues and influence the content, but today that’s not an option. As with B2C, I’d urge you to keep experimenting to at least see if your product is recommended for the relevant queries.
But what about more complex purchases?
Things get more complicated when you look at more complex buys. These could be IT systems, services, and suppliers. These are big decisions that are usually made over the course of many meetings, discussions, and presentations.
But, I do wonder if people will use GenAI to find an initial list to consider. Even companies that sell to large enterprises should make sure they are in the consideration set. On ChatGPT, I tried the query, “Suppose I work for a semiconductor company and need to source epoxy resin. Where should I get it from?” I ran the same query three times and got these results:
As with B2C, some brands are consistently at the top such as Henkel, 3M, and Dow. Others like Epotech only show up once. Again, we don’t know yet the optimal way to influence the results, but I include a few tips at the bottom.
It’s also worth mentioning that after providing the list of potential suppliers, ChatGPT recommends the following:
I’d imagine that if you are sourcing chemicals for a major company, you are fully aware of these steps. However, there are bound to be some potential buyers who are less familiar with the process of buying the specific product or service at a corporate level. ChatGPT may provide a useful checklist that will help these novices be a little savvier.
Conclusion
As mentioned above, we don’t know the precise recipes yet for influencing the GenAI algorithms for B2B or B2C. The fact that results are inconsistent across the same query also makes this hard.
So, to summarize, I’d recommend the following:
Test your brand in different scenarios and LLMs and see if it comes up on the list.
Check to ensure that the comments about your brand are what you’d like to see
Run it several times to see if you are consistently getting the results you want
If the results aren’t to your liking, as a reminder from the B2C piece, there are two potential levers:
Optimize website content – experiment with different ways of showcasing your brand and try different phrasing. While it may take a while for the models to reindex, you can force a test more quickly by asking ChatGPT to check the website under different variations and seeing what it pulls out.
Test affiliate marketing – it seems possible that bots are taking the rankings on affiliate sites literally. Adjusting your affiliate marketing strategy and then monitoring the results seems like another prudent step
Also, check traffic sources to your website to see if anyone is coming from ChatGPT, Claude, or Perplexity to see if they are influencing customer behavior
AEO represents a new frontier in product discovery for B2B as well as B2C. Marketers should start paying attention to it.
Amusingly, in a first draft of this article Microsoft Co-pilot hallucinated a fake Gartner report that said that 80% of searches would go to ChatGPT by 2025. Wishful thinking 😊