Search is changing, but the fundamentals are not disappearing. Ann Smarty, founder of Smarty Marketing and a longtime SEO expert, says AI search, generative engine optimization, and traditional SEO should not be treated as separate strategies. For manufacturers, distributors, and suppliers in residential construction, the strongest starting point is often the simplest one: make it clear who you are, what you sell, who you serve, and what problems your products solve.

In this conversation, Ann discusses how brands can build visibility across Google, ChatGPT, Gemini, Perplexity, Reddit, and other discovery channels, while avoiding common mistakes that make AI search feel more complicated than it needs to be.

Kirill Kniazev: Ann, tell me a little about Smarty Marketing and what you do.

Ann Smarty: Smarty Marketing is my first agency after more than 20 years in SEO. I founded it about three years ago, so there is still a learning curve, especially with things like leads, operations, and everything that comes with running an agency.

It probably was not the easiest time to start an agency, given everything happening in the search industry, but I am enjoying it.

Kirill Kniazev: What types of clients do you primarily work with?

Ann Smarty: We generally work with medium to large businesses. I do not love highly corporate clients because there can be so many layers between an idea and getting anything done. At the same time, we are not really built for very small businesses because so much of what we do is customized.

At this point, we work across many niches. After more than 20 years in SEO, I have seen a lot.

Kirill Kniazev: One interesting thing about your agency is that you work on both Reddit and generative AI search. Which verticals are most interested in those channels right now?

Ann Smarty: Anything connected to technology tends to be more aware of what is happening. A lot of businesses only see traffic dropping and do not know why. They do not always know that AI search, generative answers, Reddit visibility, and other changes are affecting discovery.

Technical businesses are generally more interested in generative engine optimization, or GEO. That said, I think there is still a misunderstanding around GEO. My approach is that GEO is inseparable from SEO. They need to grow together. One does not contradict the other, and most things you do for one are beneficial for the other.

When a client asks for a GEO audit, I still look at SEO because one informs the other. At this point, I do not really do separate SEO audits and GEO audits. They are very interconnected.

Kirill Kniazev: A lot of manufacturers and distributors in residential construction are starting to use more data and evaluate new opportunities like Reddit and AI search. When someone asks ChatGPT, Gemini, Perplexity, or Google AI for the best product, brand, or supplier in a trade category, what signals are those systems likely pulling from?

Ann Smarty: The fundamental signal is the training data. The model needs to know the business or the product. Even when it searches, it usually searches for branded queries around something it already knows.

So being known is probably the first signal. Even if it is a very niche problem you are solving, you need to be associated with that problem in the training data.

The second signal is organic search visibility. Once a large language model has an answer, it may search to confirm what it knows, cite something, or get more up-to-date information. LLMs search regular search engines, just like people do.

So being visible in organic search is an important step toward being visible in AI answers. Those two signals are definitely among the most important.

Kirill Kniazev: For manufacturers and distributors that already have detailed product resources, specifications, and operational information, is that a good reference point for AI tools?

Ann Smarty: Absolutely. Be as detailed as possible on your website. Also look at your content, product pages, and product feeds from a problem-solution standpoint.

That is a little different from traditional SEO. In SEO, we often cared heavily about keywords. With AI search, you need to think about the exact problems your product solves.

For example, if you sell shoes and they are waterproof, you should not only say they are waterproof. You should spell out that they are running shoes for rain. You want to move beyond keyword matching and make sure your website and product feeds clearly explain the pain points, use cases, and problems your products solve.

That kind of clarity can help elevate your brand in AI answers.

Kirill Kniazev: Many building product manufacturers and distributors have small internal marketing teams. Some use agencies, and some are doing the best they can internally. If they are limited on time and resources, what should they prioritize first?

Ann Smarty: The most untapped opportunity is their own website and existing channels. Before you invest in PR, community marketing, or anything outside your owned channels, make sure your website is very detailed about what you do, what you manufacture, what problems you solve, who your target audience is, and who you work with.

AI loves clarity. Be very clear. Avoid vague claims like “we are the best.” That is not good enough. Factual details are much more helpful. If you are a U.S.-based manufacturer, say that. If you have been in the industry for a long time, say that. Explain the products you make, the customers you serve, and the problems you solve.

You would not believe how many manufacturers and other businesses are not taking full advantage of their own websites. In many cases, just doing that well will put you ahead of the competition.

Kirill Kniazev: For manufacturers that are more seasoned and ready to branch out, Reddit can feel intimidating. It can be a wild place for large brands. As Reddit, forums, and community conversations become more visible in Google and AI platforms, how should brands participate without coming across as promotional or manipulative? Should they participate at all?

Ann Smarty: Whether they should participate is a great question. With Reddit specifically, it is sometimes better to do nothing than to do it wrong. Reddit is unique that way.

When clients come to me asking for a solution, I usually say that I need at least a month to look at their situation, read the discussions, understand the audience, and maybe test a little. It takes research.

Start by reading a lot. Understand what triggers anxiety, what triggers anger, what creates a positive response, and how to avoid causing trouble. Learn where your target audience is and how those communities behave.

If I had to name one tactic that usually works, it would be to focus on building authority before building visibility. People often give great answers, but then they cannot resist mentioning their business somewhere. As a moderator, that is often when I have to remove the answer.

So do not start by trying to drive traffic. Some subreddits do not allow links at all. Some are angrier than others. Some are more relevant, and some are less relevant but easier to engage with. Take your time to understand the landscape.

While you are doing that, build actual authority without forcing everyone to know your business name right away.

Kirill Kniazev: I appreciate that you mentioned building out the website, because there has been a lot of noise around whether websites are still relevant and whether AI is going to take away clicks. If visibility, influence, and brand inclusion matter even when someone does not click through, what should marketers measure?

Ann Smarty: That is a very good question, and we are all still figuring it out.

There are no real first-party metrics from most LLMs. Bing provides some data, which is not terrible, but it is still not very actionable. Most other measurement is still a guessing game.

I do track clients in AI visibility tools, and I look at them in comparison with competitors. But I am careful because the results are heavily affected by whatever prompts you choose. I do not focus too much on a single visibility score.

What I have started looking at more is branded search. Unfortunately, branded search impressions in Google Search Console are not perfect either, but it is still something I pay more attention to than I ever did before. I want people searching for the brand name.

I also look at brand mentions, especially on Reddit and other public spaces. Tools like BuzzSumo can measure brand name mentions over time. I want to see that growth.

For many manufacturers, the product name may be better known than the company name. Think about OpenAI versus ChatGPT. So you need to track product name mentions as well.

There may not be direct attribution in many cases now, but brand name searches, product name searches, and brand mentions are useful signals that you are doing something right.

Kirill Kniazev: It is hard to make predictions around AI, but what have you seen that gives you a sense of where things may be heading?

Ann Smarty: Honestly, most of what has happened has been disappointing from an innovation standpoint. I was looking forward to things like in-chat checkout buttons or ChatGPT building its own support for feeds instead of relying so much on Google Search. I wanted something really new.

For the past few years, the marketing industry has been talking about everything changing. But when you are doing the actual work, a lot of it still relies on Google. ChatGPT, Gemini, and other systems still search Google.

ChatGPT has done some interesting things, like app integrations and product feeds, and there were promises around working with smaller businesses to integrate product feeds. But much of that has not really materialized yet.

My hope is that these tools become more independent discovery engines so users do not have to copy and paste a product name and then go buy it through Google. I would like to see something that creates truly new behavior.

Right now, what we mostly have is a new type of audience. AI agents come to websites, interact with them in ways that are similar to humans, and search in ways that are also pretty similar. In essence, we have a new type of bot visiting our websites.

What would be genuinely new is if agents started subscribing to newsletters, reporting back to users, buying on behalf of users, or taking other actions people used to take themselves. That would be something we have not seen before.

Kirill Kniazev: That is really interesting and probably reassuring to some people. For manufacturers and distributors that want to explore Reddit or understand generative search better, how can they find you?

Ann Smarty: I always say, Google Ann Smarty. I have almost too many channels.

Smarty.Marketing has an easy way to schedule a free 15-minute conversation with me. I am also active on LinkedIn as Ann Smarty and on X/Twitter as SEO Smarty.