Online misinformation isn’t a new phenomenon, but 2020 revealed just how damaging and far-reaching its consequences could be in the midst of a global pandemic. Typically responsible for declaring a health state of emergency, the WHO also felt it was imperative to launch a webinar series on the “infodemic.” Stopping the spread of falsehoods about the virus became a component of stopping the virus itself.
Tech giants like Google, Facebook, and Twitter were forced to act as well (again, not a new conversation), turning off reviews and flagging misleading information more proactively. The Social Science Research Council proclaimed that “misinformation is everybody’s problem now.”
Going into 2021, we as consumers have a (healthy) mistrust for information sources. So the question becomes: What can an honest brand do about it? And if your business is reliant upon big tech, like Google and Instagram, to garner clicks, web traffic, and sales, how does one brand stand out from the rest?
At Yext, we write about this a lot: the more questions businesses answer on their own website — where they control the information — the better. Providing updated and transparent information prevents prospective customers from bouncing back to the third-party publishers and search engines where misinformation abounds. The best part of all? It’s entirely in your control.
But what about those third-party sites? As it turns out, there is an important pattern.
Locations that have larger Knowledge Graphs get 52% more clicks on search engines — and other 3rd party publishers
The Yext Data Insights team has found that if a Location, Healthcare Provider, ATM, etc. provides and updates more pieces of information, search engine users are more likely to trust them — in the form of clicks. The end result is these businesses see increased customer awareness, education, and activity.
To get a bit more granular: In this scatterplot, each datapoint represents one entity in our sample, for example a restaurant, health care provider, or ATM. If it had more facts stored in its Knowledge Graph — like more pictures, additional links to online ordering, or additional insurances or accepted forms of payment — it will appear further to the right on the x-axis (more clicks). Illustrated by the upward trajectory formed, we’re left with a clear takeaway: a direct relationship between the comprehensiveness of data associated with an entity and interest level on their listing. The growing red circles moving up and to the right also indicate a strong correlation among velocity of facts updated, number of total facts stored, and number of clicks.
Read on for more detailed analysis.
But here’s the thing: The biggest brands naturally have the biggest knowledge graphs. They have more items to update, and they get more clicks — simply because they have a greater digital and physical footprint (as well as, um, millions to spend on ads). So, the above scatterplot could be interpreted as a fancy way of saying big brands beat little brands at everything — and that’s not the case! So, for the purposes of this analysis, the Yext Data Insights team sought to compare entities WITHIN a business. Here’s what that looks like for one high-profile Retail brand.
To provide more context behind what we mean by size of Knowledge Graphs (the amount of information stored in the Yext Platform for a given location or other type of entity), we gave entities a profile completeness grade from “A” to “C.” The “A” group corresponds to the top 25th percentile in quantity of information stored and frequency of updates, while the “C” group corresponds to the bottom 25th percentile. We gave the entities that were “middle of the road” for that business a “B,” comprising the middle 50% in terms of Knowledge Graph size and updates. Even within this brand (where every location has the exact same name and a very similar in-store experience), more facts stored (the farther on the right of the x-axis) equals more clicks received (the higher up they appear on the y-axis). This is exemplified by the business’s Times Square location, which has both the most comprehensive knowledge graph and by far the most engagement.
Here’s another example, this time with a large insurance company. There are more outliers with this business, with strong click showings from entities in the B and C groups and weak showings from a few hard-working A group members. But, on the aggregate, the A group (averaging over 80 pieces of information stored per entity) is still garnering 45% more clicks than the C group (less than 70 pieces of information stored per entity).
Doing this at scale — looking at over a hundred thousand entities within hundreds of businesses in all major verticals to assign them Completeness Grades — shows that the entities within a business with the largest Knowledge Graphs (entities with the most facts stored, graded an A) have greater potential for getting more clicks, with both a higher “ceiling” and a lower “floor” for digital engagement.
Meanwhile, while there are always a few outliers to any good sample group, for the most part, entities with the fewest facts stored (those graded C) have both the lowest average number of clicks and seem to have their topline potential for click volume capped, compared to those with more facts stored.
To sum it all up: Across verticals, larger knowledge graphs (more facts) and more frequent updates = more engagement and clicks. Managing more business facts online isn’t a magic bullet, but paired with a robust website and an attention to detail across online platforms, it can give businesses a rare opportunity to fight brand misinformation and win more customers.
**Entities were grouped into the A cluster when they were in the top 25th percentile in terms of number of facts stored, the C cluster when they were in the bottom 25th percentile in terms of number of facts stored, and the B cluster when they were between the 25th and 75th percentile for this metric. Percentiles were assigned within businesses and within entity groups, taking the type of location into account, so as to minimize differences that can occur when comparing similar entities across different size brands, for example when comparing a local restaurant versus a restaurant that is part of a national chain or when comparing Bank Branches to ATMs. Correlation between Facts Updated, Facts Stored and Third Party Clicks, does not necessarily infer causation.