Consumer Software UX & NPS Benchmarks (2025) – MeasuringU

When we think about software, Artificial Intelligence has gotten a lot of attention lately. But throwing an AI label on a product doesn’t necessarily make people use the software frequently or more effectively.

How do we know how usable and useful a product is?

Understanding and predicting usage, adoption, and growth starts with measuring usability and usefulness.

Our research has shown that perceptions of usability and usefulness predict both intention to use and recommendation intentions. Both of these predict future usage.

To understand the current consumer software landscape, in January 2025, we conducted a large-scale retrospective benchmark with 1,896 U.S. respondents on 40 popular consumer software products, similar to what we’ve done for the last 10+ years. Respondents tended to be younger (50% below the age of 35), most (85%) had had at least some college, and a bit over half (56%) were female.

The 40 software products included a mix of popular productivity, storage, security, music streaming, language learning, and generative AI software:

1. Adobe Illustrator
2. Adobe Photoshop
3. Adobe Premiere
4. Adobe Reader
5. Amazon Music
6. Babbel
7. ChatGPT
8. Claude
9. Credit Karma
10. Dropbox
11. Duolingo
12. Firefox
13. Gemini
14. GIMP
15. Gmail
16. Google Calendar
17. Google Chrome
18. Google Docs
19. Google Drive
20. Google G Suite
21. Google Meet
22. Google Sheets
23. Google Slides
24. iCloud
25. McAfee Antivirus
26. Microsoft Edge
27. Microsoft Excel
28. Microsoft Office 365
29. Microsoft Outlook
30. Microsoft PowerPoint
31. Microsoft Word
32. Norton Antivirus
33. OneDrive
34. Pandora
35. Quicken
36. Rosetta Stone
37. Safari
38. Spotify
39. TurboTax
40. Yahoo Mail

Participants were asked to reflect on their most recent experiences with the software and answer several items, including the System Usability Scale (SUS), the UX-Lite^®, the standard likelihood-to-recommend (LTR) question used to compute the Net Promoter Score (NPS), and the Technical Activities Checklist (TAC-10™). The full details are available in the report. Here are some highlights.

The most popular method to measure loyalty is the Net Promoter Score (NPS). It’s calculated using an eleven-point (0 to 10) likelihood-to-recommend question, with the NPS computed by subtracting the percent of detractors (0–6) from the percent of promoters (9–10).

Across the 40 products, the average Net Promoter Score was 24%, ranging from −12% to 56%. Firefox (56%), Google G Suite (55%), and Google Docs (52%) had the highest scores; the lowest were for Adobe Photoshop (−9%) and Microsoft Edge (−12%).

We have collected data from ten consumer software products every few years since 2011. From 2022 to 2025 the NPS for some products declined (e.g., Microsoft PowerPoint) while others improved (e.g., Microsoft Word, Quicken).

MS Word saw one of the largest increases in Net Promoter Scores (a statistically significant difference from 2022 to 2025, Z = 2.1, p = .04). The most commonly reported reason that people like MS Word is its ease of use with advanced functionality. As one user reported:

• “I like that it has a lot of features, but for the most part, they don’t get in the way or distract you if you don’t need or want to use them.” 

Figure 1: Previous and current NPS for Microsoft Word.

We used the popular System Usability Scale (SUS) to measure the perceived usability of the 40 products. SUS is a ten-item questionnaire with possible scores ranging from 0 to 100. The average SUS score from over 500 products (including websites, consumer software, and business software) is 68. The average SUS score from this group of consumer products was 77, with a low score of 57 and a high score of 87. A raw SUS score of 77 compared to our SUS norms translates to the 80^th percentile (well above the average 50^th percentile).

The product with the lowest SUS score in this dataset was Adobe Illustrator, with a score of 56.5 (at the 22^nd percentile). The top user comments regarding issues with Illustrator were its steep learning curve and expensive pricing.

• “The GUI is extremely complicated, it’s not simple nor intuitive. It’s the one software I use that I’m basically pulling my hair out when using it to try and do SIMPLE tasks.”

The SUS scores for most of the eight products we tracked from the previous reports were reasonably stable from 2022 to 2025. Products that changed by six points or less were TurboTax, Dropbox, Microsoft Excel, Microsoft PowerPoint, McAfee Antivirus, and Adobe Photoshop. Products that changed by more than six points were Microsoft Word (eight-point increase) and Quicken (ten-point increase).

In the report, we also present findings for the UX-Lite. It’s becoming an important benchmark for many organizations as part of using a single (succinct) score to quantify software acceptance and satisfaction. It’s related to earlier research that produced the Technology Acceptance Model (TAM), but it has only one item each for ease (how easy is the software to use) and usefulness (how well do the features meet users’ needs). In aggregate, it provides a quick measure of technology acceptance (a mini-TAM).

The Impact of Ease and Usefulness on Intentions to Reuse and Recommend

We also used this new larger data set to revisit the strong relationship between perceptions of ease and usefulness to behavior intentions we demonstrated in our earlier research.

We created a new structural equation model to map the relationship between Ease and Usefulness (using the UX-lite) with behavioral intentions to continue using and to recommend products to others, mediated through Overall Experience (“How would you rate your overall experience with {product}?”; 0: Terrible, 10: Excellent) and Brand Attitude (“How would you describe your attitude toward {product}?”; 1: Very Unfavorable, 7: Very Favorable).

Figure 2: Structural equation model demonstrating the impact of perceived ease and usefulness on intentions to reuse and recommend.

The model has excellent fit statistics (c²(3) = 5.3, p = .15; CFI = 1.0; RMSEA = .02; BIC = 141.2), and the link (beta) weights are all statistically significant (p

• The combination of Ease and Usefulness accounts for over half (51%) of the variation in respondents’ ratings of Overall Experience.
• The combination of Ease, Usefulness, and Overall Experience accounts for 62% of the variation in respondents’ ratings of Brand Attitude.
• The combination of Ease, Usefulness, Overall Experience, and Brand Attitude accounts for 64% of the variation in respondents’ ratings of Likelihood to Reuse products.
• Finally, the combination of the other five variables accounts for 69% of the variation in respondents’ ratings of Likelihood to Recommend.

In summary, the model demonstrates the significant impact of perceived ease and usefulness on intentions to continue using and to recommend consumer software products, mediated by overall experience and brand attitude. This is why it is so important for companies to focus on making their products useful and easy to use.

Ease by Usefulness Scatterplot

Figure 3 shows a scatterplot of the two components of the UX-Lite scores. The dotted red lines indicate the component means from this study (78 for ease; 79 for usefulness).

Figure 3: Scatterplot of products by their UX-Lite Ease and Usefulness means (dotted lines show the overall means across products, zoomed in).

The best products landed in the upper-right quadrant. Some examples of notable positive quotes from users are:

Credit Karma

• “It allows me to dispute inaccurate information on my credit report. And alerts me of any changes on my credit report allowing me to keep up with my information.”
• “I like how user-friendly Credit Karma is for people who are just beginning to learn about their credit. I also enjoy how they give little text bubbles and notifications about certain things on your credit that you may not think of off the bat.”

Google Docs

• “The documents can be accessed from just about anything with a browser. Don’t need to download software or ensure compatibility. It’s all online so it just works everywhere.”
• “I like to collaborate on a shared document without the problem of creating multiple versions of the document with different degrees of editing.”

Pandora

• “I have been using it for years. I simply love how it can predict what I love to listen to, plus introduce me to new sounds I would most likely love to listen to.”
• “100 percent I would recommend this free service, the music is great, the quality is great. One good thing about Pandora is it’s not just music; there are a lot of comedy channels that I like to listen to. All this definitely helps pass time, especially during a boring day at home or work.”

The lower left quadrant shows the products with the lowest relative ease and usefulness scores. Some examples of notable negative quotes from users are:

Adobe Illustrator

• “Very non-intuitive design, poor cross-compatibility; for example, if you are well versed in Photoshop, much of the skills and workflow that you learn won’t work the same way in Illustrator.”
• “It has so many limitations for live drawing with certain brush types and I find that frustrating. If I draw a simple line then apply certain brushes the effect is too extreme or there is some other issue that requires me to redo certain things or adjust them, Because of this I try to limit using brushes that do not update live on the canvas unless I have a ton of time to dedicate to something I consider extremely important to get a certain look for.”

Babbel

• “The repetitive nature of some of its exercises, which can make the learning experience feel monotonous over time.”
• “They don’t offer Mandarin (Simplified Chinese) on their platform.”

Excel

• “There is definitely a learning curve if you’re doing anything moderately complex.”
• “It can sometimes be unnecessarily complicated and hard to find where to look to do the specific thing you need to get done.”

We’ve been writing about the connection between usability and for over 15 years. Figure 4 shows the SUS scores for NPS detractors, passives, and promoters (computed at the individual level for all responses; all differences are significant with p

Figure 4: SUS scores for each NPS category (with 95% confidence intervals).

This figure shows that, on average, promoters had SUS scores of 87.0 (an A+ on the Sauro-Lewis curved grading scale). The mean SUS for passives was 73.5 (a B− on the Sauro-Lewis curved grading scale) and, for detractors, was 61.2 (a grade of D). We’ve seen this pattern since we started publishing our consumer software reports in 2017. This pattern is also consistent with previous research we conducted that found that SUS scores above 80 have a good chance of indicating promoters.

Interestingly, that’s a similar relationship to what we found in our earlier analysis from 15 years ago. In both cases, promoters are associated with SUS scores above 80, and detractors fall in the 60s. That’s a consistent pattern despite the passage of time and different datasets.

Retrospective UX benchmarking is an important tool for investigating attitudes toward constructs, such as usability and usefulness, and their relationships to behavioral intentions such as the likelihood to reuse and recommend. With this information, UX researchers have estimates of how different products are doing relative to their competitors and some high-level diagnostic information that can guide additional research to understand the “why” behind the numbers.

The full list of benchmarks and analysis of verbatim responses is available for purchase in the 2025 Net Promoter & UX Benchmark Report for Consumer Software.