As the chief knowledge officer at SearchBlox Software, Dr. John Lewis's expertise is on the topics of human capital and strategic change within the knowledge-driven enterprise.
SearchBlox Software is a sponsor of CMSWire's DX Summit Fall Event, taking place online on Oct. 28 and Oct. 29. Lewis will be presenting a session on “Optimizing the Search Journey for Insights and ROI” on Oct. 29.
We spoke with Lewis to learn more about the SearchBlox approach to optimizing the search journey.
Where AI Fits in Search Optimization
SMG: Tell us how AI works in improving search optimization. What are the metrics and information sources that go into play?
Lewis: The first thing that comes into play with AI is a search strategy. Then we feed that strategy with different information sources and data, as well as specific AI capabilities where they align most in the strategy. For example, we have a strategy around search setup that is different than the strategy for executing search. During setup, AI improves search optimization by automatically generating better document titles while the content is being indexed. This solves a long-standing problem within search where the search engine is doing its part in connecting a search query to terms in a document, especially the title, yet users see poor results due to poor document titles. This problem is solved with AI that is good at generalizing a title from the content in the document. Then we use specific AI to optimize each of the four stages of our search strategy, which are before typing, during typing, after typing and after selecting. Without affecting privacy, a persona can still be constructed so that certain search terms and patterns feed the AI to be able to provide personalized results for that specific persona.
SMG: Has AI for search optimization replaced the role of web architects and content managers in manually improving SEO? How does/can/should the role of writers shift when orgs implement this technology?
Lewis: We think these roles will still be needed, and we can support them by providing insights or by automatically creating better metadata for each document. Writing content well is a different skill set than knowing how to make a document findable, by many users that may use different terms in their searches. So, we certainly aren’t thinking in terms of replacing content managers, but we can help them do their job better.
We see clients who have content that goes back 10 or 15 years, and the documents have really bad titles and descriptions. It would be too time-consuming and expensive to have content managers go back and make updates to all of the older content to help the documents become more findable in search. This is where AI can step in and automatically perform a task that just isn’t feasible with manual labor. So, this is an example of how we see this technology helping and not replacing content managers.
Seeing the Results of Search Optimization
SMG: What improvements have you seen in ecommerce websites when search is optimized, and how do you correlate the improvement metrics with personalization? Is it a direct or assumed correlation?
Lewis: Optimizing search used to be a topic mainly about search relevancy ranking and filters to narrow the search results. But viewing and selecting from the list of search results is considered stage three of our four stage model of search. Stage one and stage two have made the biggest difference in ecommerce, providing most of the benefit, where you are providing suggestions before and while they are entering their search query. So, you can guide people to popular products and promotions that they didn’t know existed and would not have searched for.
When we implemented this strategy on an ecommerce site, we immediately started seeing an improvement in conversions, even two times the previous numbers. So the improvement numbers come from direct data and not just assumptions. And search optimizations are still happening at stage 3 as well, where we can tailor search relevance settings to specific types of content. And then stage 4 is all about optimizing the entire search journey, looking at the insights on how different personas move through their search journey, with different choices and different results, and this feeds back into the continuous improvement loop. So, we don’t view a search installation as a static implementation of technology, but more of a lens into living documents which provides direct benefit but also value from the overall insights towards continuous improvement.
SMG: As it relates to search, do you have some general advice on product value that you could share?
Lewis: When we started, implementing search was hard and expensive. Implementations could take around a year where now it is about a month. Some vendors would charge one dollar per document to index, which becomes very expensive for large companies. So, our value proposition, when we started, was asking if we could provide a more accessible, simple to use search, that still includes all the expected capabilities, to provide the best value to customers.
And we are still holding to the same value proposition today. Some key advice that our CEO, Robert Selvaraj frequently shares is that just because a product is really expensive does not mean that it is really good. The key to success is in providing more value at less cost, and the key to sustainable growth is not going into long projects without finding value as quickly as possible. Providing faster time to value is the goal, because it combines the idea of adding value as well as the idea of speed in the realization of that value.
SMG: Tell us a little about your session topic, “Optimizing the Search Journey for Insights and ROI.”
Lewis: In watching people search for information, usually we don’t see one-and-done behaviors. Even when using voice search at home, people may ask if it will rain that day, but usually there are a series of questions, not a single question. People are not just looking for a single answer, they are usually making decisions and looking for more details and explanations to be able to make the best decision. This takes them on a search journey, as a better way to describe the pattern of behaviors. So, the quest has been about creating an experience with search technologies that better facilitates this journey.
In optimizing the search journey for insights and ROI, this session looks at how insights lead to continuous improvements, and how this leads to a return on investment. Knowing the top search terms is interesting, but knowing more about the personas using search, and the content they are seeking, is more insightful, and more actionable as feedback in a continuous improvement process.
Claim your free pass to DXS here and be sure to check out the session, ““Optimizing the Search Journey for Insights and ROI.”