gshc2020.com

If your brand becomes a common answer in “small business banking,” does that make it more likely to show up in “business credit cards” or “small business loans”?

The short answer: it depends on whether you’re measuring citations or brand mentions. In the brands we analyzed, citations tended to spread more readily across categories than brand mentions did. And the brand’s expertise and depth in a category shaped how far its authority appeared to travel.

Key takeaways

• Among the brands we analyzed, a domain could be cited in almost any category if its content met the kind of source criteria ChatGPT relies on (relevance, authority, freshness). Getting mentioned by name in ChatGPT tended to happen mostly in categories close to a brand’s core expertise.
• For brand mentions specifically, depth in a category tended to matter more than breadth. Brands that consistently appeared across a category’s prompts in ChatGPT were the ones associated with successful expansion into new categories. Showing up once in 20 categories didn’t appear to build a higher brand-mention share.
• Industry shaped the pattern. In the brands we analyzed, finance and real estate saw broader coverage associated with more citations. Legal and healthcare set a higher bar: even full prompt coverage and repeat citations didn’t reliably translate into brand-mention share.
• Topical authority tended to give brands a stronger starting point when expanding into adjacent categories, not a free pass into unrelated ones. Expansion worked best when the next category was genuinely close to one the brand already answered well. 

Methodology

We partnered with Kevin Indig and Growth Memo again to analyze data from the Semrush AI Visibility Toolkit.

The dataset covers 1,094 US categories tracked monthly in ChatGPT from January through June 2026, with five prompt variants per category.

It includes 283,215 domain-category citation observations and 76,493 brand-mention observations across January through May, plus 45,578 category-expansion appearances by 1,458 mapped brand entities.

Definitions

Two signals are tracked separately throughout the study:

• Citations are the sources ChatGPT links to as evidence in its answers
• Brand mentions are the brand names ChatGPT includes directly in the response text (which may be neutral, positive, or negative)

For the relatedness analysis, a brand’s “core expertise” is defined as any category where the brand had already appeared in at least three of five prompts before the target-month appearance.

“Category closeness” is calculated using semantic similarity between the target category’s prompts and the brand’s prior expert categories.

1. The bar for being mentioned is higher than the bar for being cited

When a brand in our study appeared in a topical category in ChatGPT, whether it was just cited or also mentioned by name depended heavily on how related the category was to its core expertise.

In categories close to the brand's core expertise, brands were cited as a source in 74% of appearances, mentioned in 44%, and both cited and mentioned in 34%. In categories distant from the brand's core expertise, those shares dropped to 50% cited, 25% mentioned and 9% both cited and mentioned.

Being cited as a source held up reasonably well across the distance (50% distant vs. 74% close). But being named alongside a citation was strongly tied to relatedness — cited-and-named appearances were nearly 4x higher in close categories than in distant ones.

The pattern was even sharper if you looked only at citations. Of the citations that happened in the least related categories, just 18% also included a named brand. In the most relevant, 46% did. 

Related categories were more than twice as likely to turn a citation into a named-brand mention in this sample.

The takeaway: match your strategy to your goal. Publishing broadly can build citation presence, and that appears to work even in categories far from your core. But if you want ChatGPT to name your brand when a customer asks a question, focusing your effort on categories close to what you’re already known for tends to pay off more.

2. Depth matters more than breadth in ChatGPT for brand recognition

In our study, a brand could show up once in 20 categories and own none of them. Depth in one category was associated with a higher brand-mention share. Breadth across many categories, without depth, was not.

Citation share, however, responded to coverage either way. A brand that showed up in even one of a category’s five prompts tended to gain some citation share. A brand that showed up in all five gained more. In the sample, there was no obvious spread-thin penalty for citations.

Brand mentions behaved differently. A brand that appeared in only one of a category’s five prompts was associated with a drop in mention share. That penalty didn’t disappear until the brand reached at least three of five prompts in a category. Only after that did breadth stop hurting brand mentions.

“Being cited in many categories does not show a spread-thin effect in this sample. But brand recognition works differently. The categories where a brand consistently earns named mentions are the ones from which it can credibly expand.”

Kevin Indig, Founder of Growth Memo

The takeaway: similarly, prioritize depth in a few categories over presence in many. The data doesn’t support the idea that a site can serve unlimited categories, but it also doesn’t support a universal cap. What it does support is a simple rule: earn repeat presence in a category before treating it as a launching pad for the next one.

3. Industry shapes how far authority spreads in ChatGPT

According to our data, how coverage breadth affects a brand’s presence in ChatGPT isn’t the same across industries.

In finance and real estate, wider coverage tended to hold more citation leverage. A brand that showed up in all five prompts in a category was associated with stronger citation share than one that showed up in only one.

That's not to say any finance or real estate brand can dominate every adjacent category. It just means that in these industries, spreading coverage across many categories didn't appear to hurt a brand's citation share.

Legal and healthcare set a higher bar. Their citation gains from broader coverage were weaker than in finance or real estate. And for brand mentions, coverage didn’t translate: even at full five-prompt coverage, mention share still trended negative in those fields. Being widely cited didn’t translate into being named.

How to plan topical category expansion in AI search

To expand successfully in AI search, don't chase every content topic. Start with the areas where you're already recognized, then expand outward in small, deliberate steps, focusing on what moves the needle for your brand.

Here’s how to plan an expansion move:

1. Map where you already have authority. List the categories where your brand consistently earns both citations and named mentions across most of the related category prompts. Those are your anchor categories, not every category where you’ve appeared once.

Start by checking the topics you already appear for in Semrush’s Visibility Overview tool:

1. Pick target categories that are genuinely close to your core. In this study, closeness was measured by semantic similarity between the anchor and the target category. In practice, that often means looking for categories where the customer, problem, product, and buying decision look similar and are directly relevant to your business.

Open the Topic Opportunities tab to see what you should focus on next:

1. Focus on what earns brand mentions. To earn mentions in a new category, focus on clear brand positioning across your content, third-party coverage from credible sources (Reddit, review sites, industry publications, expert commentary), and evidence of expertise (author credentials, primary research, real product details). These weren’t tested directly in this study, but they are common signals that tend to impact AI visibility.

An avid reader since childhood, Pradershika's idea of extreme sports is staying up to read “just one more chapter.” She lives in India.

Areas of Expertise

Read Full Bio

August 3, 2026

We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.

Credit: Woot/Lifehacker

Table of Contents

Anker’s floatable speaker, the Soundcore Boom 2, is down to $69.99 at Woot, a substantial discount from its usual $129.99 price and $40 less than Amazon’s current price of $109.99—making this its lowest recorded price, according to price trackers. Plus, shipping is free for Prime members ($6 for everyone else). This deal is scheduled to run for three weeks, although it could end sooner if the speaker sells out.

It has an IPX7 rating and can float, so it's fine around pools or at the beach without needing constant attention. If it falls into the water, it will come back up on its own—your music will cut out while it is submerged, since Bluetooth cannot transmit through water—but in most situations, the bigger advantage is that the speaker itself survives without any issues.

Sound-wise, the Boom 2 delivers up to 80 watts through a setup that includes two tweeters and a dedicated woofer, which is enough to fill a large room or carry across an outdoor space. There’s also a BassUp feature that boosts the low end; turning it off noticeably reduces the impact, so you will likely want to keep it on for a fuller, heavier sound. You can also pair a second unit for stereo sound, though it is limited to Bluetooth since there’s no wifi support. Speaking of, battery life is rated at up to 24 hours, but that figure assumes moderate volume with lights and bass boost turned off. In real use, especially at higher volumes, expect less.

What do you think so far?

The design focuses more on durability than refinement, and the sound, while loud, does not have the same balance you get from speakers like the JBL Flip 6 or Sony SRS-XB23, especially in the mids and highs. The companion app helps by letting you adjust the EQ and control the lighting, which gives you some flexibility to shape the sound. Still, this is not meant for detailed listening sessions or for someone who prioritizes audio precision, notes this PCMag review. It works best for people who want a speaker that can handle unpredictable environments, play loud without much effort, and keep going through a full day of use.

Our Best Editor-Vetted Tech Deals Right Now

Deals are selected by our commerce team

Lifehacker has been a go-to source of tech help and life advice since 2005. Our mission is to offer reliable tech help and credible, practical, science-based life advice to help you live better.

• Our Mission
• Our Team
• Newsletter

© 2001-2026 Ziff Davis, LLC., A ZIFF DAVIS COMPANY. ALL RIGHTS RESERVED.

Lifehacker is a federally registered trademark of Ziff Davis and may not be used by third parties without explicit permission. The display of third-party trademarks and trade names on this site does not necessarily indicate any affiliation or the endorsement of Lifehacker. If you click an affiliate link and buy a product or service, we may be paid a fee by that merchant. %!s()

Both of these additions are to Google’s Optimize Your Crawl Budget Documentation that is for enterprise websites with over a million pages and medium sized websites of over 10,000 web pages that rapidly change. Although the documentation is not aimed at every website, the information on that page is useful for understanding how crawling works.

Google’s Crawlers Share A Website’s Crawl Capacity

Google added the following passage:

“While each crawler has a different crawl demand, the crawl capacity limit is shared across all crawlers. This means that high demand from one crawler can reduce the capacity available for others.”

This means that if a site gets crawled by Googlebot-Image and Googlebot, both of the crawlers are sharing the crawl budget for the website. The impact is that a high rate of crawling by one bot can reduce the capacity available to other Google crawlers.

Google Recommends 304 Server Response

The other interesting recommendation that Google added was the suggestion to use 304 Not Modified server response code. According to the Mozilla Developer Network web documentation:

“The HTTP 304 Not Modified redirection response status code indicates that there is no need to retransmit the requested resources.”

The thing about the 304 Not Modified response code is that there is no redirection happening, even though technically all 3xx server response codes are classed as redirection codes. Despite that technical detail, nothing is being redirected.

Google’s updated documentation says:

“Use HTTP caching: Support 304 (Not Modified) HTTP status codes. If a page hasn’t changed since Google last crawled it, returning a 304 code tells Google to reuse the cached version, saving your server bandwidth and resources.”

A 304 Not Modified response tells Google’s crawlers that a page has not changed since the last time it was crawled. The result is that the server doesn’t serve the page to Googlebot and Googlebot can go focus on indexing other pages.

Read Google’s updated documentation here. The two changes described here are the only new additions. There are other changes that here and there that make Google’s documentation more precise, which is documented here.

SEJ STAFF Roger Montti Owner - Martinibuster.com at Martinibuster.com

I have 25 years hands-on experience in SEO, evolving along with the search engines by keeping up with the latest ... %!s()

Lunar Ice Could Support Future Moon Missions

The findings were published in Science Advances on July 31, 2026, as space agencies prepare for a new era of lunar exploration. NASA's Artemis program is aiming for crewed missions to the moon's south polar region in 2028, where permanently shadowed craters may contain valuable deposits of water ice.

For astronauts, that ice could serve several purposes. Once melted and purified, it could provide drinking water. Electricity could also separate it into oxygen for breathing and hydrogen for rocket fuel. A reliable local supply would reduce the amount of material that future missions need to transport from Earth.

"It's crucial to identify any materials on the moon that an astronaut can make use of while they're up there," said Nicholas Schmerr, an associate professor in UMD's Department of Geological, Environmental, and Planetary Sciences and a co-author of the study. "Since they will be limited by the few resources they brought from Earth, anything they find on the moon will help them basically live off the land, especially for longer-term missions or outposts."

Satellites Cannot See Deep Underground

Scientists still do not know exactly how much ice exists on the moon or where most of it is located. Orbiting satellites can examine the lunar surface, but their instruments mainly detect material in the uppermost layer of soil.

Some of the moon's water ice may be buried much farther below the surface. The new study suggests that seismic measurements could reveal deposits that orbital observations cannot reach.

Frozen and dry lunar soil respond differently when seismic waves move through them. Ice makes the surrounding material stiffer, allowing vibrations to travel two to three times faster than they do through dry soil.

Ice-rich areas can also reflect seismic energy instead of allowing it to continue through the ground. The effect is similar to sound bouncing off a wall and producing an echo. According to Schmerr, a seismometer placed in the right location on the moon could detect both of these changes.

"We can use seismic waves to not just see whether ice is present but also roughly how much of it there is," he explained.

Testing the Seismic Signature of Moon Ice

The research team examined the idea in three different ways.

Lead author Harrison Lisabeth (Ph.D. '16, geology), a rock physicist at Lawrence Berkeley National Laboratory and UMD alum, worked with volcanic rock from Arizona. When crushed, the material closely resembles lunar dust. Lisabeth froze the rock and used X-rays to observe how ice formed inside the tiny spaces between individual grains.

Co-author Matthew Siegler of the University of Hawaii created detailed temperature models for the moon's south polar region. These maps helped identify craters that have remained cold enough to preserve ice for billions of years.

At UMD, Schmerr used computer simulations to model small moonquakes traveling through underground deposits of lunar ice. Across all three approaches, the presence of ice produced distinct and measurable changes in the seismic results.

Lunar Ice May Preserve Ancient Solar System History

Moon ice could be useful for astronauts, but it may also contain important evidence about the early solar system.

Deeply shadowed lunar craters can trap water and other volatile materials for extremely long periods. Because the surrounding rocks are about four billion years old, the ice preserved there may offer clues about how water moved through the young solar system.

"The moon witnessed some of the most critical parts of the early solar system, including how water was delivered," Schmerr said. "Studying the ice deposited there could reveal how water spread and ultimately how Earth's oceans formed."

Upcoming Missions Could Test the Method

The researchers may soon have an opportunity to compare their predictions with real measurements from the moon.

China's Chang'e-7 mission is expected to land near Shackleton Crater in late 2026. The mission will carry a seismometer, and several suspected ice deposits are located nearby.

NASA's Artemis astronauts could also deploy the Lunar Environmental Monitoring Station in 2028. Schmerr helped develop the instrument for seismic exploration.

"Our findings are laying the groundwork for an observation we'll get in the next couple of years," Schmerr said. "No one has physically measured the ice on the moon yet, but we now have a prediction for what to look out for. That's an important first step."

This research was funded by the U.S. Department of Energy Office of Science, Office of Basic Energy Sciences, Chemical Sciences, Geosciences, and Biosciences Division (Contract No. DEAC02-05CH11231), and the NASA Solar System Exploration Research Virtual Institute CLEVER project (Grant No. GR00024738) and GEODES project (Grant No. 80NSSC19M0216). This article does not necessarily reflect the views of these organizations. %!s()

In America, bigger is almost always better. We have bigger dinner plates than do most of the rest of the world. We also have bigger refrigerators to hold all that food needed to fill those bigger plates. That necessitates a larger kitchen to house the big fridge, demanding a bigger house around that, and likely a comparatively large driveway leading to that big-ass house. That super-sizing could make any normal-sized car or truck parked on the property seem dinky by comparison. 

In such a case, the mega-big six-door, lifted, and wide-tired 2006 Ford F350 we looked at last Friday would be the perfect accompaniment to any such oversized McMansion. Unfortunately for its builder/seller, a $55,000 asking price proved too large for most of our tastes. That was obvious in the comments and the extremely overwhelming 90% 'No Dice' loss the fancy Ford suffered in the vote.

The Rotary Club

For the longest time, the Ford Motor Company maintained a relationship with Japan's Mazda Motors. Ford sold some massaged Mazdas as Mercury models, and shared its popular Explorer—albeit only in two-door form—with their Asian partners. At no time, however, did Ford ever sell any of Mazda's Wankel-engined products as its own. In fact, aside from NSU and a dalliance by Citroën that actually drove the company to bankruptcy, Mazda has pretty much gone it alone in cheerleading the rotary engine.

Perhaps the most famous and beloved model to carry the triangle piston torch is the RX-7, and today we're looking at a fairly unmolested 120,000-mile 1978 GS edition to see what all the fuss is about.

This Spark Yellow over beige vinyl RX-7 hails from the model's first year on the market. It entered the U.S. market replacing the Cosmo, becoming Mazda's sole rotary offering in North America over three ensuing models and its RX-8 replacement.

They don't build 'em like they used to

According to the seller, they have owned the car for 13 years, and it came with a fresh build of its 12A two-rotor motor at that time. That 1146cc Wankel put out a laudable 100 horsepower when new. Per the ad, this one has a Racing Beat header and exhaust, so perhaps it makes a couple more ponies.

Everything looks stock under the sloping hood, although a missing belt on what looks to be an air-injection pump does raise questions. Backing the engine up is a five-speed manual transmission driving the solid rear axle. Braking is by discs in the front and drums in the rear, masked by handsome factory four-spoke alloy wheels.

Those wheels look to be in great condition, as does the bodywork atop them. Age has taken its toll on the matte black paint on the windscreen pillars and license plate lamp housings, but aside from those minor issues, it's difficult to tell the car from new. Unfortunately, the ad doesn't say whether it drives as well as it looks.

Where there's wear, there's bound to be tear

One thing that can be gleaned from the pictures in the ad is that the car's interior has not staved off Father Time as well as the exterior has. 1970s plastics don't have the same mojo as modern ones do, a fact made obvious by the cracked dashboard, broken door lever surrounds, and split seam in the driver's seat. Additionally, there are faded window cranks, worn silver paint on the dash separator, and a tear in the passenger-side door cover.

Do any of those issues kill this Mazda's appeal? Probably not. Taken as a whole, though, the interior's flaws should offer some wiggle room in negotiating a deal on the car. Replacement parts for these cars are going to be next to impossible to obtain, after all, so its current condition is likely going to have to be lived with by its next owner.

One final note about the interior is the stereo, as it's in stereo, with a more modern unit having been placed in the dash under the factory-installed one. That's kind of weird. In the plus column, this is a GS model and so comes with a sunroof. That's a glass panel, which the seller claims is a fairly rare inclusion.

A humdinger of a deal?

Aside from the stereo and the exhaust, this SA RX-7 appears to be completely stock and in very decent condition. It also carries a clean title and will come with some original literature and a few maintenance receipts. That's everything you might expect from a well-cared-for car.

To take possession of this RX-7, there is the small matter of the $11,799 the seller is asking for it. That, they warn, is the only deal they are willing to entertain, as trades are off the table. Additionally, they only want to deal with serious inquiries, so low-ballers and shenanigan-eers need not apply.

What do you think about this tidy if flawed RX-7 when its $11,799 asking price is taken into account? Does that seem fair for the car as it's presented in its ad? Or do the interior foibles demand a lower asking price?

You decide!

Chicago, Illinois, Craigslist, or go here if the ad disappears.

Help me out with Nice Price or No Dice. Hit me up at [email protected] and send a fixed-price tip. Remember to include your commenter handle. %!s()