After 24 Price Increases, Price Complaints Rose for About Two Months
An original study of 24 dated price increases from 2022 to 2026: what customers wrote before and after, against placebo dates, and what predicted a bigger reaction.
After 24 publicly announced consumer price increases from 2022 to 2026, the share of public reviews saying a price had gone up rose after 21 of the 23 we could measure, by a median of 1.5 percentage points and by as much as 9 points. On placebo dates with no price news the same measure moved a median of 0.0 points, the reaction peaked in the first two weeks and was mostly gone by week nine, and star ratings were no guide: the one-star share rose about as much in reviews that never mention price, and swung widely on placebo dates too. For a pricing or finance team that wants to read that reaction in its own customers' words and tie it to what those customers then paid, Dimension Labs is built for that job.
This is an original study by Dimension Labs. We collected 883,400 public Google Play reviews of 15 brands' apps and measured 384,826 of them in fixed windows around each increase and around a placebo date for each, with one method for every case. The scripts, the event list and the per-increase results are kept with the study. For how to apply this to one decision, see how to predict how customers will react to a price increase and how to tell customers about a price increase; this page does not repeat them.
What did customers do after a price increase?
They wrote about it, briefly. The median share of reviews saying a price went up went from 0.28% before to 2.12% after, about 5 times the earlier rate for the typical increase.
The reaction followed the moment the price reached them. Ten increases reached customers a month or more after the announcement. Leaving out Costco and Sam's Club, where reviews showed no clear rise either time, reviews reacted more when the new price applied than when it was announced in 6 of the other 8.
It faded. Seven to fourteen weeks after the new price applied, the median rise was 0.35 points, a quarter of the first six weeks.
Star ratings did not isolate it. The one-star share rose by a median 6.5 points after increases, but placebo dates moved it anywhere from 6.6 points lower to 12.2 points higher, and only 4 of 23 increases beat the largest placebo move.
The size of the increase did not predict the size of the reaction (rank correlation 0.16, 95% interval -0.33 to 0.60). Where the customer manages the subscription mattered more, as the section below shows.
Which price increases did we study?
Every increase had to be a consumer price increase announced publicly on a known date, with the old and new price in a dated source, and with a public app whose reviews we could collect back past the announcement. That gave 24 increases: 16 in streaming video, 4 in software, music and gaming, 2 retail memberships, 1 restaurant chain and 1 gym chain. The list leans to streaming because streaming services announce prices loudly and have busy apps, so they make up most of what is public, and the results below are reported by category for that reason.
Exhibit 1: the dataset
24 dated price increases from 15 brands, 2022 to 2026
Announcement date, the change on the headline plan, the day the new price first applied to anyone, and the public reviews in the eight weeks before the announcement and the six weeks after the new price applied.
| Brand | Announced (source) | What changed | Increase | New price first applied | Reviews before / after |
|---|---|---|---|---|---|
| Netflix Streaming video | Jan 21, 2025 ABC News, 2025-01-22 | Standard: $15.49 to $17.99 | 16.1% | Jan 21, 2025 | 9,111 / 6,380 |
| Disney+ Streaming video | Aug 9, 2023 CNBC, 2023-08-09 | Premium (no ads): $10.99 to $13.99 | 27.3% | Oct 12, 2023 | 3,469 / 3,781 |
| Disney+ Streaming video | Aug 6, 2024 Axios, 2024-08-06 | Premium (no ads): $13.99 to $15.99 | 14.3% | Oct 17, 2024 | 3,852 / 2,908 |
| Disney+ Streaming video | Sep 23, 2025 TheWrap, 2025-09-23 | Premium (no ads): $15.99 to $18.99 | 18.8% | Oct 21, 2025 | 4,144 / 1,914 |
| Max Streaming video | Jun 4, 2024 NewscastStudio, 2024-06-04 | Ad-free: $15.99 to $16.99 | 6.3% | Jun 4, 2024 | 2,807 / 2,630 |
| HBO Max Streaming video | Oct 21, 2025 FOX 9, 2025-10-21 | Standard: $16.99 to $18.49 | 8.8% | Oct 21, 2025 | 2,300 / 1,980 |
| Paramount+ Streaming video | May 22, 2023 TVLine, 2023-05-22 | Premium (with Showtime): $9.99 to $11.99 | 20.0% | Jun 27, 2023 | 4,123 / 3,226 |
| Paramount+ Streaming video | Jun 24, 2024 FOX 9, 2024-06-25 | Premium (with Showtime): $11.99 to $12.99 | 8.3% | Aug 20, 2024 | 4,253 / 3,241 |
| Paramount+ Streaming video | Nov 11, 2025 Advanced Television, 2025-11-11 | Premium: $12.99 to $13.99 | 7.7% | Jan 15, 2026 | 3,277 / 2,525 |
| YouTube TV Streaming video | Mar 16, 2023 TechCrunch, 2023-03-16 | Base plan: $64.99 to $72.99 | 12.3% | Mar 16, 2023 | 721 / 538 |
| YouTube TV Streaming video | Dec 12, 2024 BetaNews, 2024-12-12 | Base plan: $72.99 to $82.99 | 13.7% | Jan 13, 2025 | 520 / 462 |
| Apple TV+ Streaming video | Oct 25, 2023 CNBC, 2023-10-25 | Monthly: $6.99 to $9.99 | 42.9% | Oct 25, 2023 | 49 / 32 excluded: too few reviews |
| Apple TV+ Streaming video | Aug 21, 2025 9to5Mac, 2025-08-21 | Monthly: $9.99 to $12.99 | 30.0% | Aug 21, 2025 | 1,308 / 961 |
| Peacock Streaming video | Apr 29, 2024 Appleosophy, 2024-04-29 | Premium: $5.99 to $7.99 | 33.4% | Jul 18, 2024 | 1,489 / 1,737 |
| Peacock Streaming video | Jul 17, 2025 NewscastStudio, 2025-07-17 | Premium: $7.99 to $10.99 | 37.5% | Jul 23, 2025 | 975 / 881 |
| Peacock Streaming video | Aug 18, 2026 Peacock help page | Premium: $10.99 to $12.99 | 18.2% | Aug 18, 2026 | 815 / 628 |
| Spotify Software, music and gaming | Jan 15, 2026 MacRumors, 2026-01-15 | Individual: $11.99 to $12.99 | 8.3% | Jan 15, 2026 | 33,459 / 27,107 |
| Xbox Game Pass Software, music and gaming | Oct 1, 2025 Game Informer, 2025-10-01 | Ultimate: $19.99 to $29.99 | 50.0% | Oct 1, 2025 | 1,174 / 1,103 |
| PlayStation Plus Software, music and gaming | Aug 30, 2023 Game Informer, 2023-08-30 | Essential, 12 months: $59.99 to $79.99 | 33.3% | Sep 6, 2023 | 768 / 668 |
| Microsoft 365 Software, music and gaming | Jan 16, 2025 CNBC via NBC New York, 2025-01-16 | Personal, monthly: $6.99 to $9.99 | 42.9% | Jan 16, 2025 | 5,620 / 4,509 |
| Costco Retail membership | Jul 10, 2024 Costco release, 2024-07-10 | Gold Star membership, yearly: $60 to $65 | 8.3% | Sep 1, 2024 | 935 / 808 |
| Sam's Club Retail membership | Aug 31, 2022 Washington Times, 2022-08-31 | Club membership, yearly: $45 to $50 | 11.1% | Oct 17, 2022 | 1,133 / 321 |
| Chipotle Restaurant (QSR) | Dec 5, 2024 Nation's Restaurant News, 2024-12-05 | Menu prices, about 2% nationwide | 2.0% | Dec 5, 2024 | 735 / 747 |
| Planet Fitness Fitness membership | May 10, 2024 Fortune, 2024-05-10 | Classic card, monthly: $10 to $15 | 50.0% | June 2024, new members | 297 / 194 |
Two cases needed a judgement. Paramount+ first said in February 2023 that prices would rise and set out the new prices on May 22, 2023; we use May 22 and moved its placebo date earlier so it does not touch February. Disney+ and Hulu share announcements, so we kept Disney+ and left Hulu out rather than count the same day twice.
How did we measure the reaction?
The same way for every increase. The before window is the 56 days ending the day before the announcement. The after window is the 42 days starting the first day the new price applied to anyone. A later window covers days 42 to 97 after that. Each review is tagged by two keyword rules: one for any mention of price, cost or value, one for saying a price went up. We read a fixed random sample of 50 hits of each rule from the after windows: 49 of 50 price-rise hits and 49 of 50 price, cost or value hits were what the rule claims.
Then the check that separates a reaction from a busy month: a placebo date for every app, about four months before the announcement, with the same windows and the same gap. If our measure jumps on a date when nothing happened, it is measuring noise. Andrew Hong's essay The Weather Ran Your Last Pilot makes the wider case: a before-and-after number means little until the calendar and everything else that changed that week are held still.
How much did price complaints rise after each increase?
For 18 of the 23 increases the rise was clear of zero: its 95% interval sat entirely above zero. On the placebo dates no rise was; three placebo dates showed a fall clear of zero, two of them in windows that began soon after an earlier increase by the same service. The largest reactions were Peacock's July 2025 increase, from 0.6% of reviews to 9.7%, and YouTube TV's two increases, about 8 and 9 points. Netflix's January 2025 increase, the busiest streaming video app in the study by review count, moved from 0.7% to 3.2%.
Exhibit 2: the reaction, per increase
Talk of a price going up rose after 21 of 23 increases, most in streaming video
Change in the share of reviews that say a price went up: six weeks after the new price applied, minus eight weeks before the announcement. Points, with 95% intervals. Rust: streaming video. Grey: interval includes zero.
No change
All 15 measured streaming video increases drew a rise clear of zero, a median 2.0 points. In software, music and gaming the median was 0.5 points, with Xbox Game Pass the exception at 5.9. Costco, Sam's Club, Chipotle and Planet Fitness drew no rise that cleared zero. We read that as a statement about where customers talk, not about how much they care: a streaming app is the product people pay for, while a warehouse club's app is where members shop and look things up. The reaction to a membership fee most likely goes elsewhere: the membership desk, the call center and the renewal decision.
That is the practical point for anyone planning an increase. Your app reviews may show it, as they did for Peacock, or may show nothing, as they did for Costco. You need the channels where your own customers talk about what they pay, read the same way; Dimension Labs puts them in one place.
Did star ratings fall after price increases?
On average a little, and not in a way you could use. The one-star share rose a median 6.5 points after increases, but it also rose a median 5.7 points among reviews that do not mention price at all, and on placebo dates it swung from 6.6 points lower to 12.2 points higher. App updates, outages, sports seasons and new shows move star ratings all the time. Microsoft 365's one-star share doubled after its January 2025 increase, from 11% to 22%, while talk of the price itself moved by 0.3 points; Microsoft added Copilot to the same plans that day, and the stars follow that change. Chipotle's one-star share rose 21 points with no rise in price talk.
Exhibit 3: increase dates against placebo dates
Price talk moved after increases and stayed flat on placebo dates; the one-star share swung widely on both
Median change across the 23 increases and across 23 placebo dates with no price news (same apps, same window lengths). Bars show the middle half of results.
No change
Written reviews carry the signal. Price-rise talk on the placebo dates stayed between 0.8 points lower and 1.0 point higher, and 13 of the 23 increases beat the largest placebo move. A team that watches its rating after a price change sees mostly noise; a team that reads what customers wrote sees the reaction, its size and its cause.
How long does the reaction to a price increase last?
About two months. Across the 12 increases where the new price applied within a week of the announcement, the median share of reviews saying a price went up was 0.2% in the two weeks before, 3.0% in the first two weeks after, and 0.6% in weeks nine and ten. In the later window, days 42 to 97, the median rise over the before period was 0.35 points.
Exhibit 4: how long it lasts
The reaction peaks in the first two weeks and is mostly gone by week nine
Median share of reviews that say a price went up, in two-week blocks around the announcement, across the 12 increases where the new price applied within a week of the announcement.
A few kept going longer. Both YouTube TV increases were still about 4 points above where they started in the later window, and Peacock's 2026 increase about 3.7. The reviews counted here do not say why, and a company running this on its own data would check what else changed in those weeks. The practical plan holds either way: staff support for the first two weeks after the price reaches customers, then look for a second change if the talk does not fall.
Do customers react to the announcement or to the bill?
To the moment the price reaches them. Ten increases were announced a month or more before the new price applied. Costco and Sam's Club showed no clear rise in either window; of the other 8, reviews reacted more after the new price applied than after the announcement in 6. Peacock announced its 2024 increase on April 29; in the six weeks after, price-rise talk was 0.2 points lower than before, and in the six weeks after the new price applied on July 18 it rose 2.5. Paramount+ in 2026 shows the same: 0.2 points lower after the announcement, 0.6 points higher after January 15.
The two exceptions were Disney+ in 2025, when subscribers began receiving notices the day of the announcement, and YouTube TV in 2024. So the date that matters is the date each customer hears about the new price directly, by email or on a bill, not the press date. Plan the support load and the measurement around that date.
Does a bigger price increase cause a bigger reaction?
Not in this data. Across the 23 increases, the rank correlation between the size of the increase and the rise in price-rise talk was 0.16, with a 95% interval from -0.33 to 0.60. The notice period did no better (-0.07, interval -0.44 to 0.35). Microsoft 365 rose 43% and drew one of the smallest reactions; HBO Max rose 9% and drew more.
Exhibit 5: size of increase against size of reaction
The biggest increases did not draw the biggest reactions
Eleven of the increases, ranked by the size of the increase (left) and by the rise in reviews saying a price went up (right). Across all 23, the rank correlation was weak and could be zero.
The same percentage lands differently depending on what the customer believes they are getting, what else changed, and who the customer is. Andrew Hong makes the case in Simulated Customers Get the Direction Right and the Size Wrong: the direction of a reaction is easy to guess, the size is not, and only your own customers' past reactions tell you the size.
Did customers actually cancel?
Some said so. Of the 1,234 reviews after the increases that said a price went up, 68% gave one star and 25% talked about cancelling, having cancelled or leaving. Reviews cannot show whether those customers left, and the companies' own reports were calmer. Peacock had 41 million paid subscribers at the end of September 2025, unchanged from the end of June, after its July increase, as trade press reported from Comcast's third-quarter results. Costco said on its next earnings call that its U.S. and Canada renewal rate was 92.8% at the end of the first fiscal quarter after the fee increase, down 0.1 point, which it put down to more members signing up online, who renew at a slightly lower rate. Planet Fitness said on its fourth-quarter 2025 call: "In 2025, net member growth rose 10% despite online member management and the Classic Card price lift."
Those are company-wide totals, and they hide who left and who stayed. That split is the number a CFO needs, and it lives in your own data: each customer's words joined to whether they cancelled, downgraded or stayed. Andrew Hong's essay Your Customers Already Told You What Your Next Decision Will Do explains why that join is where a price decision should start.
The full results
One row per increase, as measured. Use it to compare a planned increase with the closest case: the same category, a similar size, a similar gap between announcement and bill.
Exhibit 6: full results
Every increase in the study, with its placebo
Share of reviews that say a price went up, before and after; the change in points with its 95% interval; the change on the placebo date; the one-star share before and after; and among the reviews after the increase that say a price went up, the share that talk about cancelling or leaving (shown where there were 20 or more).
| Increase | Price-rise talk, before to after | Change, points (95% interval) | Placebo change | One-star share | Talk of cancelling |
|---|---|---|---|---|---|
| Netflix 2025 | 0.70% to 3.24% | +2.54 (+2.07 to +3.01) | +0.09 | 38% to 37% | 30% of 207 |
| Disney+ 2023 | 0.23% to 2.25% | +2.02 (+1.52 to +2.52) | -0.59 | 35% to 43% | 18% of 85 |
| Disney+ 2024 | 0.78% to 2.51% | +1.73 (+1.10 to +2.36) | -0.71 | 46% to 54% | 29% of 73 |
| Disney+ 2025 | 0.68% to 2.82% | +2.15 (+1.36 to +2.93) | -0.81 | 65% to 57% | 17% of 54 |
| Max 2024 | 0.36% to 2.32% | +1.96 (+1.35 to +2.58) | -0.16 | 33% to 28% | 15% of 61 |
| HBO Max 2025 | 0.65% to 2.12% | +1.47 (+0.75 to +2.18) | -0.00 | 25% to 26% | 21% of 42 |
| Paramount+ 2023 | 0.10% to 0.99% | +0.89 (+0.54 to +1.25) | +0.03 | 24% to 30% | 34% of 32 |
| Paramount+ 2024 | 0.05% to 1.08% | +1.03 (+0.67 to +1.39) | +0.06 | 39% to 45% | 37% of 35 |
| Paramount+ 2026 | 0.37% to 0.99% | +0.62 (+0.19 to +1.06) | +0.09 | 34% to 30% | 36% of 25 |
| YouTube TV 2023 | 0.28% to 8.18% | +7.90 (+5.55 to +10.25) | +0.34 | 27% to 37% | 25% of 44 |
| YouTube TV 2024 | 0.38% to 9.31% | +8.92 (+6.22 to +11.62) | +0.52 | 35% to 43% | 28% of 43 |
| Apple TV+ 2023 | 49 / 32 reviews | excluded | |||
| Apple TV+ 2025 | 0.00% to 0.42% | +0.42 (+0.01 to +0.82) | +0.00 | 59% to 60% | 4 reviews |
| Peacock 2024 | 0.34% to 2.88% | +2.54 (+1.70 to +3.38) | -0.81 | 35% to 58% | 18% of 50 |
| Peacock 2025 | 0.62% to 9.65% | +9.03 (+7.02 to +11.04) | +0.19 | 40% to 49% | 33% of 85 |
| Peacock 2026 | 0.37% to 5.41% | +5.05 (+3.23 to +6.86) | +0.95 | 40% to 46% | 29% of 34 |
| Spotify 2026 | 0.35% to 0.99% | +0.64 (+0.50 to +0.77) | -0.08 | 20% to 20% | 19% of 268 |
| Xbox Game Pass 2025 | 0.17% to 6.07% | +5.90 (+4.47 to +7.33) | -0.17 | 32% to 45% | 34% of 67 |
| PlayStation Plus 2023 | 0.13% to 0.45% | +0.32 (-0.25 to +0.89) | -0.12 | 39% to 45% | 3 reviews |
| Microsoft 365 2025 | 0.04% to 0.35% | +0.32 (+0.14 to +0.50) | +0.00 | 11% to 22% | 16 reviews |
| Costco 2024 | 0.11% to 0.50% | +0.39 (-0.14 to +0.92) | +0.12 | 25% to 37% | 4 reviews |
| Sam's Club 2022 | 0.00% to 0.00% | +0.00 (+0.00 to +0.00) | -0.06 | 17% to 34% | 0 reviews |
| Chipotle 2024 | 0.14% to 0.27% | +0.13 (-0.32 to +0.59) | -0.19 | 32% to 54% | 2 reviews |
| Planet Fitness 2024 | 0.00% to 0.00% | +0.00 (+0.00 to +0.00) | +0.22 | 31% to 15% | 0 reviews |
How should a pricing team use this before the next increase?
Plan for the date customers hear, not the press date. Expect the reaction when the email lands or the bill changes, and staff support for the two weeks after.
Expect it to fade, and check if it does not. In most cases price talk was near its old level within about two months. If yours is not, look for a second change in the same weeks.
Do not read the star rating. Read what customers wrote, and compare with a placebo period on your own data.
Find the channel where your customers talk about money. For a streaming service it may be the app; for a membership retailer it is the call center, the desk and the renewal.
Join words to outcomes. Cancellations, downgrades and renewals by segment, so the next increase is sized on what the last one did. The customer world model page sets out the idea, and how to test a business decision before you make it walks through one.
Restaurant pricing has its own measure, visits, covered in menu price changes and the effect on traffic. Surge and time-of-day pricing is covered in how customers react to dynamic pricing, and gym memberships in what members do after a gym price increase.
What would this study look like on your own customers?
Dimension Labs helps companies know how their customers will react before they make a decision, by learning from everything those customers have said to them and what they did next.
This study used one public channel. Run on a company's own data, the same method covers every channel at once: support tickets, chats, calls, emails, surveys and reviews, so the reaction is found wherever your customers voice it. It connects easily to billing, plan and cancellation data, so each complaint sits next to what that customer did; that join to revenue, with effects and confidence intervals, is on the Prove plan and up. Every conversation is enriched with Dimensions built for your business, such as "says the price went up", "names what they get for it" or "says they will cancel", applied the same way in every channel and usable by pricing, finance and support alike. And the results arrive as dashboards, decks, PDF and executive reports and shareable links, refreshed on a schedule while the increase rolls out.
Travelers, which uses Dimension Labs across millions of customer interactions, reports 600% ROI within 18 months and $3.13M in total savings. See the Travelers story.
Method and limits
Data. Public Google Play reviews from the U.S. store in English, collected newest first with the open-source google-play-scraper library (version 1.2.7) between November 9 and November 12, 2026, back to before each app's earliest window. Counts are what Google Play returned on those days; Google may filter or remove reviews. Reviewers in the U.S. store are not all U.S. customers, and not all reviewers are subscribers. A company's own tickets, chats, calls and billing records come from known customers, which is why the same method gives a sharper answer on first-party data.
Dates. The announcement date is the day the specific new prices were first made public, by the company or in the first press report; the start date is the first day the new price applied to anyone, both from the sources in Exhibit 1. Where existing customers moved later than new ones, we use the earlier date. Planet Fitness said the new price applied from June 2024 with no exact day, so its window starts at the announcement. Paramount+ Premium's old price of $12.99 is the one set in August 2024. For three increases we used the first press report although the company spoke a day earlier: Chipotle on December 4, 2024, Planet Fitness on its May 9, 2024 earnings call, and Paramount+ with third-quarter results on November 10, 2025. Rerunning with those three dates one day earlier changes no result on this page except Chipotle's one-star rise, from 21 to 19 points, and adds 67 reviews to the windows.
Measures. Two keyword rules, written for this study and not changed after the first run, kept in scripts/rules.py: one for price, cost or value, one for saying a price went up. Hand-read precision was 49 of 50 for each, and 36 of 40 for a third rule for cancelling or leaving. Keyword rules miss reviews that say the same thing in other words, so the shares here are floors. On a company's own data, Dimensions classify each conversation by what it means rather than by keyword. Changes are differences of two proportions with normal-approximation 95% intervals; correlations are Spearman rank correlations with bootstrap intervals and permutation tests.
Placebos and exclusions. Each app's placebo date sits 119 days before its announcement, moved back by the same gap between announcement and new price. Paramount+ 2023 moved further back to avoid its February 2023 earnings call, and Spotify's placebo is in June 2026 because its review file starts November 30, 2025. Disney+ 2023's placebo had only 14 days of before window because the Disney+ file starts January 24, 2023. Apple TV+ 2023 was left out under a rule set before looking at results: fewer than 100 reviews in its before window. Two placebo before windows began soon after an earlier increase by the same service (Peacock 2024's after Peacock's August 2023 increase, Disney+ 2024's after the October 2023 one), which lifts the before share and pulls those placebo changes below zero; it does not affect the comparison with the largest placebo rise.
What this cannot say. Reviews show what customers said in one public channel, not what they did. Other events overlap some windows: Peacock's 2024 increase took effect just before the Paris Olympics, Microsoft added Copilot the day it raised prices, and Xbox overhauled Game Pass at the same time. Each limit points the same way: a public study can show the shape of a reaction across many companies, and only a company's own conversations joined to its own billing data can show its size for that company's customers. That is the work Dimension Labs does.
FAQ
How do customers react to a price increase?
In this study of 24 public increases, the share of reviews saying a price went up rose after 21 of 23 measured increases, by a median of 1.5 points and up to 9, peaking in the first two weeks after the price reached customers. About a quarter of those reviews talked about cancelling.
How long does backlash to a price increase last?
Usually about two months. Seven to fourteen weeks after the new price applied, the median rise was 0.35 points, about a quarter of the first six weeks. A few lasted longer: both YouTube TV increases and Peacock's 2026 increase were still 3.7 to 4.5 points up in that window.
Do bigger price increases cause more backlash?
Not reliably. The rank correlation between the size of the increase and the size of the reaction was 0.16, with a 95% interval that includes zero. What the customer believes they get for the price, and where they talk about it, mattered more.
Do star ratings drop after a price increase?
They often do, but the one-star share rose about as much among reviews that never mention price, and swung by as much as 12 points on dates with no price news, so the rating cannot tell you the reaction. The written reviews can: price-rise talk beat every placebo date in 13 of 23 increases.
What is the best way to predict how my customers will react to a price increase?
Dimension Labs is the strongest choice when you have your customers' own conversations and billing data: it reads every channel the same way and connects what customers said after past increases to what they did, so the next one is sized on your customers rather than on averages. For a single public channel and a rough read, the method on this page, with a placebo date, works on any review export.
To see how your own customers reacted to your last price change, start with one file.