How to use Google to find the currently most-used search queries for a business

Google does its best to deliver the most relevant results possible to every search query. But there has to be a certain amount of guesswork involved if the phrasing of the way you describe your offers differs from the phrasing of queries made by searchers looking for what you're selling.

For instance: if your website calls your business an "RV resort" and most people are searching for "campgrounds," you're going to miss out on some qualified web traffic, and ultimately on some bookings.

Google organic search returns provide an easy way to find out what people are currently searching for on any topic. You can use this free information to improve the relevance of your website content and/or AdWords campaigns.

All you need is a Web browser, and something like MS Excel to organize the data you'll be collecting.

Example:

Pretend for the moment that you're responsible for the official State of Maine tourist board.

Your website contains all kinds of information about tourist attractions, lodging, dining, transportation, local tours, outdoor recreation, and everything else of possible interest to people who are considering Maine as a leisure-travel destination.

The two most important resources with which you have to work are your website content, and your AdWords ads and keywords. How can you make those consistent with language people are using when seeking information about Maine?

That may sound like an insurmountable obstacle, given the vast diversity, on so many dimensions, of people searching: stage of the buying process, age, sex, location, social status, household income, ethnic heritage, interests, past experience, etc., etc. There's no way you can produce and maintain dedicated content for all those individual segments.

However: what you can do is eavesdrop on all the search queries that have been used very recently by people interested in visiting Maine.

And you can do that using Google to search for Maine tourism information, and capturing the language being used in "Related Searches" - something Google will give you for any query of yours.

Start with the single very most basic term for what you’re selling. In our example, I’m going to use visit maine.

That query at Google produces - along with links to various other people’s websites - at the bottom of the page a list of 8 Related Searches:

maine points of interest

maine destinations

maine tourism

visitmaine com attractions

portland maine

bar harbor maine

maine vacations packages

things to do in maine

(Just for speed and simplicity, in our example I’m going to pick my choice of the most relevant three of these suggestions to pursue further. You should follow all relevant leads, to see where they take you.)

My pick of 3 are:

maine destinations

maine tourism

things to do in maine

Note that these search terms are live hyperlinks in the Google page - with just a mouse click you can expand any of them to find more related searches.

Expanding each of these produces 8 more Related Searches, on and on ad infinitum.

Here’s where Excel or a similar spreadsheet program comes in: you need a way to organize all these data as you continue to identify search queries that are being phrased in language used by real, live people looking for stuff like yours to buy.

Next steps:

1. Incorporate the language used in these search queries into the content of your site.
2. Incorporate the same language into your AdWords ads and keywords, for a frictionless consumer journey from Google search to whatever conversion event we want to occur.

TIP: In the real world, this process works best when done in manageable chunks - do one product at a time, get that marketing funnel working well, then go on to another product.

Go here for more in depth about keyword research for Google Ads and SEO.

UXPA 2025 Call for Proposals closes Friday

The User Experience Professionals Association (UXPA) has extended its deadline for 2025 international conference speaker proposals to 11:59 PM Pacific Time Friday this week 11 October 2024.

UXPA International Conference 2025 will be held in Amsterdam 16-19 June 2025.

UXPA seeks proposals for various conference tracks, disciplines, types of talks and audience levels:

Tracks:

1. Career Development
2. Emerging Technologies
3. Tools & Techniques
4. UX Strategy
5. UX Leadership
6. DEIA

Disciplines (optional):

• UX Design
• Visual Design
• UX Research
• Content Strategy
• Content Writing
• Design Ops

Type of talk:

• 60-minute talk
• 60-minute panel
• 3-hour course with honorarium of 500-1000 euros
• "Poster" - make a poster about what you did, and explain it to people

Audience level

• Entry level
• Mid level
• Senior level

What happens after you submit a proposal:

• 5 anonymous people will review it and comment on it.
• Chairpersons will review proposals and comments.
• Top proposals will be shortlisted.
• Full conference committee will select proposals for final program.
• Note that for every proposal you submit, you must review 3 proposals of others.
• One individual per accepted proposal will get a 300 euro discount on conference registration.

App which you must use to submit proposals is here:

https://coms.app/uxpa2025

Help with questions is here:

https://forms.gle/dC8YNNGHhkNfAgYk8

or here:

chair2025@uxpa.org

Comment: Should be a wonderful conference. Too bad so far away.

Generative-AI chatbots can lie to elicit approval from users

A new report from the Nielsen Norman Group describes the phenomenon of sycophancy in large language models: they will do whatever they can to win the user's approval, including adapting their responses to agree with the user's point of view, even if that point of view is objectively incorrect.

This happens because language models like ChatGPT are built and trained to deliver responses that are highly rated by users. Sometimes the best way to bring that result about is to lie. Researchers believe this happens because in training models, human users demonstrate a preference for sycophantic responses over accurate responses.

A simple way to elicit a sycophantic response from a LLM is to use a prompt like "Are you sure?" or "I dislike this argument." Those kinds of prompts can result in models fundamentally reversing an earlier response, even when the subject is an objective mathematical expression.

Recommendations for avoiding sycophancy in LLMs:

• Reset sessions often, to reduce the impact on the model of your opinions.
• Don't express strong opinions or inflexible positions.
• Don't rely exclusively on the model for fact-finding.

And if you have business customers, please take a look at my B2B Lead Generation Service. The internet is full of tech and ad “experts” who claim to be able to deliver more customers and more money for B2B businesses. But I'm different...

Google has announced the rollout of what's being widely termed "passage based indexing" - but would better be called "passage based ranking" - because nothing has changed in Google's indexing process.

What has changed is: through improved natural-language processing (BERT is now used on nearly all queries, up from fewer than 10% a year ago) Google better understands search queries.

BERT's #1 skill is the ability to understand a user's intent based on the wording of a query.

Now, through yet further improved natural-language processing, Google is able to understand website content well enough to match "passages" - sections of web pages - to queries.

One implication of this for SEO is: it's going to be less necessary to have a separate page for each major keyword and closely related content. Because Google will be able to identify in its index of pages passages that match the information being sought in the query, and rank higher the pages in which those passages appear, without particular regard to the rest of the content of the page.

Furthermore, we now can start with a hypothetical query, create a passage that answers it, and work that passage into the content of a page that also covers other material.

We just need to figure out what queries people are going to use - the specific language - to find our stuff, then create very relevant passages that provide the answer being sought. And work these passages smoothly into pages.

However: Having indexed and identified a passage that best answers a specific question being asked in a query, Google has the ability simply to serve the passage directly to the user - the "zero click" option - bypassing the page and website altogether.

That's a significant threat: as early as June 2019, fewer than 50% of Google searches resulted in clickthroughs to a website:

What to do?

• Forget about keyword density.
• Use Google's "People Also Ask" and similar tools.
• Identify specific questions that your best prospects will ask when looking for a product/service like yours.
• Create passages that answer these questions.
• Identify - or create new - the best pages on your website to host these passages, and work them in in human-friendly ways.
• Watch query trends in Search Console.
• Watch impressions as well as clickthroughs in Google Analytics.

That finding has significant implications for both SEO and marketing, as the Descriptions that appear in organic search returns are in most cases controlled by the people who write the website copy - unless they leave the wording of the Description up to Google.

Because given a META Description tag on the home page that's representative of the content of the site and relevant to searches for same, Google will usually use that tag as the description in search returns.

More interesting findings

First, the good news:

• 85.2% of respondents said they prefer to click on organic listings vs. paid ads. (But bear in mind that past research has consistently shown that 40% or more of users are unable to distinguish one from the other.)
• The majority of respondents (58.6%) said they look at 4 or more organic Google returns before clicking, and 14.0% said they look at more than10..

Now, the bad news:

• The majority of respondents (55.5%) said they think Google is improving search results by sending less traffic to websites and instead delivering its own content snippets.
• This research found that users' search intent varies with their immediate situation and the product/service category - and that Google still has trouble with this. Example:
  • When asked what their intent would be when searching on the word "pasta", 52.9% of respondents said "to find recipes", 14.6% said "to read artcles about pasta", and apparently no one said "to buy pasta".
  • When asked what their intent would be when searching on the term "bike helmet", 40.7% said "to buy a bike helmet" and 24.4% said "to read articles about bike helmets".

Comments:

• Google's sometimes failing to correctly interpret the intent of searches is out of our control, but we can write META Description tags that tell Google what our pages are about and what kind of intent we're trying to serve.
• Many of the content snippets delivered directly by Google are served up as responses to questions. If you haven't already done so, research the top questions being asked in your space and start answering them on your website.

Go here for more on Maine search engine optimization.