# The short version

A location's profile is good or bad only next to the businesses a customer sees beside it, and that comparison tracked real Google customer actions. The best-standing tenth of a brand's locations drew about **185 actions for every 100 an average location drew**, and the worst about 60. The comparison also moves, because nearby competitors keep adding photos, reviews, and updates, and a weak profile is not stuck, since the poorest locations that put in work over six months grew while the ones that coasted fell.

**What to do:** Watch who appears when a nearby customer searches your category, and count where you beat them and where you trail.

## How to read the win rate

Several charts use a **win rate**, the share of about 73 operational signals, such as photo count, review count, rating, and freshness, on which a location beats the businesses next to it in local search. Because a strong brand and a weak brand carry different baselines, every brand's locations are split into ten equal groups from worst to best, called **deciles**, so a top performer and a bottom performer are always read against their own brand.

- **D10**: the best-standing tenth of a brand's locations  
- **D6**: about the average location  
- **D1**: the worst-standing tenth

## Finding 01

### The target is whoever is _next door_

Advice about local profiles used to arrive as a fixed number. Get onto 75 percent of the long-tail directories, fill 90 percent of your core fields, answer 85 percent of your reviews. A location would reach the number, mark it done, and move on. The trouble is that a fixed number says nothing about the businesses a customer actually sees beside you, and those businesses do not hold still. While one location rested, its neighbors added photos, gathered reviews, and refreshed their hours, so the same profile that cleared the bar in October was behind by April without a single field changing.

So this study set aside fixed targets and asked a local question instead. In your own neighborhood, do you have more photos than the businesses around you, more reviews, a higher rating, fresher updates? The answer defines a location as strong or weak against its real competitors rather than against a national rule of thumb, and it is a moving answer, recomputed as the field around each location changes.

We are not claiming a single win-rate number tells a location what to fix on a Monday morning. It is a coarse read, closer to a smoke alarm than a repair plan, useful for telling a strong profile from a weak one at a glance. The specific work always comes back to the exact signal a nearby competitor is beating you on.

## Finding 02

### Better standing tracks more _actions_

The point of a profile is what customers do when they find it, so the study measured Google customer actions rather than search rank, which swings day to day. To compare a small brand against a large one fairly, each location's actions were scored against the average location in its own brand, in standard deviations. Read across the deciles, the line rises. An average location sits around the sixth decile, and the distance from the worst tenth to the best tenth is about 1.37 standard deviations of customer actions.

Indexed customer actions by profile standing, within brand  
Yext Research · Fig 1 · summer 2026  
 
| Profile Standing | Customer Actions |  
|----------------|------------------|  
| Worst decile D1 | 100                |  
| Average D6     | 185                |  
| Best decile D10 | 185               |

Actions are indexed so the average location scores 100. The best-standing tenth of a brand's locations draws about **185** and the worst about **60**, so the best profiles pull in roughly three times the actions of the worst, after adjusting for brand size and industry.

A single well-drawn slope could be one lucky brand, so the study rebuilt it brand by brand, keeping the 483 brands with at least 200 locations. A positive relationship between standing and actions appeared in 476 of them.

- **98.6%**: The slope held in **476 of 483 large brands** with 200 or more locations, and across all 13 industries.

## Finding 03

### Why the pattern holds

These are hypotheses rather than settled conclusions, and each one can be checked against your own market.

- **Beating peers is what a customer sees.** A profile is judged next to the others in the same result, so more photos or better reviews than the businesses beside you is the comparison a customer makes before choosing.
- **Standing is steadier than rank.** Rank can turn on the hour of the day, while a fuller, fresher, better-rated profile keeps drawing actions across weeks.
- **A complete profile signals an open, active business.** Recent photos, answered reviews, and current hours read as a business that is open and running.
- **The comparison measures work, not scale.** Because each location is scored against its own brand, a small location that beats its local field outscores a flagship that coasts.

## Finding 04

### Each signal has its own _shape_

Rolling every signal into one win rate is useful for a glance, but the individual signals do not all behave the same way against a location's neighbors. Three of the most visible ones show three different curves, which is why matching a static maximum on all of them wastes effort.

How three signals reward beating your neighbors  
Yext Research · Fig 2 · shapes observed across the panel · summer 2026

| Signal      | Behavior  |  
|-------------|-----------|  
| Review count| steady climb |  
| Photo count | calm middle, steep ends |  
| Google attributes | rises, then a ceiling |

Each panel plots how a location does on customer actions as it moves from far behind its neighbors on the left to far ahead on the right.  
Build reviews wherever you can, and do not be the location with half the photos of everyone nearby. Fill attributes to the local norm and stop adding past it.

## Finding 05

### AI rewards the same _standing_

The same profiles that draw more customer actions also fare better in AI answers. Splitting 310 brands into five win-rate groups, the study read how often each group's locations were mentioned in AI answers and how often they were spoken about positively, across four models from October 2025 through March 2026. Both rise with standing, though this is a cross-brand correlation rather than proof of cause.

AI mention and positive-sentiment rates by win-rate group  
Yext Research · Fig 3 · 310 brands · 4 models · summer 2026

| Win-rate group | Mention rate | Positive-sentiment rate |  
|----------------|--------------|------------------------|  
| Q1 lowest      | 42.3%       | 74.8%                  |  
| Q2             | -            | -                      |  
| Q3             | -            | -                      |  
| Q4             | -            | -                      |  
| Q5 highest     | 53.9%       | 85.7%                  |

Moving from the lowest win-rate group to the highest, the **mention rate** climbs from 42.3 percent to 53.9 percent, and the **positive-sentiment rate** from 74.8 percent to 85.7 percent, each about eleven points. The middle groups wobble, as thinner slices do, but the ends are clearly apart.

## Finding 06

### Poor profiles can _climb_ back

The harder question is whether a weak profile is stuck. To answer it the study took each brand's poorest quarter of locations at the October scan, then measured their Google actions over the year before that scan and the year before the April one, so seasonal swings cancel out. Sorting those locations by how much their win rate moved over the six months, the ones that improved grew their actions and the ones that declined fell.

Median yearly change in customer actions, weakest locations  
Yext Research · Fig 4 · summer 2026

| Change  | Customer actions |  
|---------|-----------------|  
| Declined | -3%           |  
| Big improvement | +1.5%     |

In the typical case, shown here as the median, those that let their win rate slip fell about **3 percent** in actions year over year, while those that improved the most grew about **1.5 percent**. That is a swing of about **5 points** between coasting and working, with a p-value near 0.0005 behind it, so the gap is very unlikely to be noise.

Customer actions are calls, direction requests, and clicks to a site, so at a steady conversion rate a five-point gap in actions is a five-point gap in the business those actions feed, repeated each year. The median is the cautious read. On the mean, which a few strong movers pull upward, improvers grew about 5 percent against about 1 percent for decliners, and the gap reaches 6.5 points in healthcare.

**85%**: The best-profile locations draw about **85 percent more actions** than an average location. The five-point yearly swing is one year of movement against that gap.

One caution on reading this is that the comparison sets improvers against decliners rather than running a controlled test, so it shows a strong association and stops short of proving the work by itself caused the growth.

## Finding 07

### The turnaround pays off by _industry_

The turnaround story holds everywhere, but the payoff is larger in some industries than others. The gap in yearly action growth between the big improvers and the decliners among weak locations runs from about three points in food and hospitality to more than six in healthcare, so the same effort buys more in a field where a full, current profile is scarce.

Action-growth advantage of improvers over decliners, by industry  
Yext Research · Fig 5 · summer 2026

| Industry        | Advantage |  
|----------------|-----------|  
| Healthcare     | +6.5%    |  
| Retail         | +5.2%    |  
| Financial services| +4.1% |  
| Food & hospitality | +3.4% |

The top two fields in those turnarounds were accuracy, whether the same details are synced and consistent everywhere a location appears, and visibility, whether the location is present in as many of the right places as it should be. Completeness and freshness came next. One field ran higher than expected, the change in a location's rating over the past year, which mattered almost as much as the rating itself.

## What you can do

### Three practical steps

01

#### Benchmark against your block  
Look at the businesses that appear beside you in local search and count where you lead and where you trail on photos, reviews, rating, and freshness.

02

#### Recheck as the field moves  
A profile that met a goal last quarter can be behind now, so read the comparison again rather than trusting a target you hit once.

03

#### Fix the signal you trail on  
Build reviews where you can, match the local norm on photos and attributes, and keep details accurate and current, because the weak profiles that climbed were the ones that put in that work.

Know your competition, beat your competition, and keep improving. **Read your own field first.** A full audit can come later.

## Common questions

### Questions readers ask

1. **Do better local profiles perform better?**  
   Yes. Across 500 brands and 600,000 locations, a location's standing against nearby competitors tracked its Google customer actions. The best-standing tenth of a brand's locations drew about 185 actions for every 100 an average location drew, and the worst about 60. The pattern held in 476 of 483 large brands.

2. **What is a metric win rate?**  
   It is the share of about 73 operational signals, such as photos, reviews, ratings, and freshness, on which a location beats the businesses that appear next to it in local search. It is a coarse read of standing, not a to-do list.

3. **Are poor profiles doomed to stay behind?**  
   No. Among each brand's weakest quarter of locations, those that raised their win rate over six months grew their customer actions while those that let it slip fell, a swing of about five points year over year, with a p-value near 0.0005.

4. **Does profile quality affect AI answers?**  
   It tracks with them. Across 310 brands split into five win-rate groups, the highest group was mentioned in AI answers about 54 percent of the time against 42 percent for the lowest, and was spoken about positively about 86 percent of the time against 75 percent. This is a correlation, not proof of cause.

5. **Why not just hit a fixed profile target?**  
   Because a fixed target ignores what nearby competitors are doing, and they keep adding photos, reviews, and updates. A location that met a static goal can fall behind without changing anything because the field around it moved.

## Methodology

### How this was measured

The panel: 500 brands across 13 industries, covering about 600,000 physical locations. Each location was audited for at least six consecutive months and had been live on Google for at least twenty, so both a current profile and a history exist.

The win rate: For each location, 73 operational signals (photo count, review count, average rating, attributes, review responses, freshness, and more) were compared with the businesses appearing beside it in local search. The win rate is the share of those signals it beats its neighbors on. Locations are split into within-brand deciles so strong and weak brands are read on their own scales.

The outcome: Google customer actions, scored in standard deviations against the average location in the same brand, so a large brand and a small one can be compared without size driving the result.

Repeatability: The standing-to-actions relationship was rebuilt for the 483 brands with 200 or more locations. A positive slope appeared in 476 of them, and across all 13 industries.

The turnaround test: Each brand's weakest quarter of locations at the October 2025 scan was tracked by its Google actions over the year before that scan and the year before the April 2026 scan, so seasonality cancels. Improvers and decliners were compared by median and by winsorized mean, with the median reported here.

The AI cut: 310 brands split into five win-rate groups of about 62 each, read across four AI models from October 2025 through March 2026, for mention rate and positive-sentiment rate. Cross-brand correlation, not causal.

Limitations: The links are correlational. Customer actions are a Google measure, not total sales. The single win rate is deliberately coarse, and which signal matters most varies by industry and by place, a question for follow-up work.