Customer KPI Library: 53 Scorecard Measures

You owe it to your customers to review these 53 important key performance indicators

Key Takeaways

Use the list as a menu, not a checklist. Measuring all 53 isn't the point; see what likeminded organizations track.
Ask what behavior each measure drives before you adopt it. Keep the KPI only if that is the behavior you want.
Track both driver and outcome measures. Both matter, so don't get stuck sorting each KPI into one category.
Compare customer lifetime value with acquisition cost. A ratio below one means a customer costs more to win than the profit they bring.
Compare candidates with other lists, such as these KPI examples. Then settle on the critical few that fit your strategy.

Google AI answers

Get our answers first in Google.

Henry Ford once said, “It is not the employer who pays the wages. Employers only handle the money. It is the customer who pays the wages.”

If you subscribe to the same line of thought as Mr. Ford, then you agree that the customer is of paramount importance. But if you’re not tracking the right measures, you simply can’t be sure that your customers will continue to buy your products or use your services and thus, “pay your wages.”

The best way to combat this issue is to track key performance indicators (KPIs) that are in line with your organizational strategy. Customer KPIs are often divided into two categories: drivers, and outcomes. Driver measures are used to understand if your customers are buying your products or services, and outcome measures are used to help you determine if your marketing and product development efforts are driving sales. (Don’t get too hung up on whether your KPIs are outcome or driver measures, because both are important!)

‍Our suggestion is to think about what kind of behavior each measure is driving and whether it is the kind of behavior you’re looking for. Marketing automation has made gathering and storing customer data simple, so the possibilities for data-driven measures are quite expansive. To get you on the right track, we’ve documented 53 customer KPIs and scorecard measures that you may want to consider implementing.

‍Note: We’re not suggesting nor advocating for you to begin measuring all of these KPIs. Rather, you can use this extensive list to get an idea of what likeminded organizations may be looking at, research KPIs from other lists, and then decide upon the critical few KPIs that are in line with your unique strategy.

Life Cycle KPIs

Conversion Rate

LeadingBetter when higherHelps determine the success of a particular customer interaction by tracking the percentage of interactions that result in a sale.

Formula: (Interactions with Completed Transactions) / (Total Sales Interactions) = (Conversion Rate)
Example: 4,000 ÷ 160,000 = 2.5%.
What good looks like: e-commerce conversion typically runs 2–3%; well-optimized stores clear 4%+
Weak → strong: "Sell more online" → "Raise site conversion from 2.5% to 3.1% by Q3, owned by the ecommerce lead via checkout redesign".

Cross-Selling Rate

LaggingBetter when higherMeasures a brand's ability to sell a consumer a product related to the one they have purchased, thereby expanding wallet share.

Formula: (Customers with an upsell or cross-sell ÷ total existing customers) × 100.
Example: 120 ÷ 1,000 = 12%.
What good looks like: there is no universal number — it depends on how many products you sell, so grow it against your own baseline and watch that expanded customers stay and renew
Weak → strong: "Sell more to existing customers" → "Raise the upsell/cross-sell rate from 12% to 15% this fiscal year, owned by the account management lead via quarterly business reviews".

This rate can be measured in additional products sold or additional revenue gained from these products.

Customer Churn Rate

LaggingBetter when lowerIndicates the percentage of customers that either fail to make a repeat purchase or discontinue their service during a given period.

Formula: (Number of Customers Lost in a Given Period) / (Number of Customers at the Start of the Period) = (Customer Churn Rate)
Example: 50 lost ÷ 1,000 = 5%.
What good looks like: lower is better, but it varies sharply by industry and contract model — compare with sector peers, separate voluntary from failed-payment churn, and drive the trend down
Weak → strong: "Lose fewer customers" → "Cut annual churn from 5% to 4% by year-end, owned by the VP of Customer Success, reviewed monthly".

Show all 14 Life Cycle KPIs

Customer Lifetime Value

LaggingBetter when higherThe net profit a company anticipates gaining from a consumer over the entire length of the buying relationship.

Formula: Average order value × purchase frequency × customer lifespan × gross margin.
Example: $85 × 4 orders/yr × 3 yrs × 50% = $510.
What good looks like: there is no universal number — read it against customer acquisition cost, and grow it through repeat purchases and retention rather than discounts that cut margin
Weak → strong: "Make customers worth more" → "Raise CLV from $510 to $560 within a year, owned by the CRM lead via a loyalty program that lifts orders from 4 to 4.4 a year".

This can help to determine the costs and benefits of acquisition efforts.

Customer Lifetime Value / Customer Acquisition Cost

LaggingBetter when in target rangeThe ratio of a customer's lifetime value to the cost to acquire a customer.

Formula: (Net Expected Lifetime Profit From the Customer) / (Cost to Acquire the Customer) = (Customer Lifetime Value / Customer Acquisition Cost)
Example: $18,000 ÷ $3,000 = 6.0:1.
What good looks like: 3:1 is the widely cited healthy target — below 1:1 you lose money on every customer, while above ~5:1 can mean you're under-investing in growth
Weak → strong: "Our customers are worth more than they cost to win" → "Hold blended LTV:CAC at or above 3:1 while scaling spend 20%, reviewed monthly by RevOps."

This ratio should ideally be greater than one, as a customer is not profitable if the cost to acquire is greater than the profit they will bring to a company.

Customer Profitability Score

LaggingDetermines which customers make the most and least substantial contributions to profit.

Formula: Revenue from a customer − total cost to serve that customer.
Example: $50,000 − $38,000 = $12,000.
What good looks like: there is no universal number — rank customers by profit after cost to serve, then grow the most profitable group and reprice or restructure the accounts that lose money
Weak → strong: "Focus on profitable customers" → "Raise average profit per key account from $12,000 to $15,000 within one year, owned by the account management director via cost-to-serve review".

Companies can use this to determine where they will allocate spending and investments.

Customer Retention Rate

LaggingBetter when higherMeasures the portion of consumers who remain customers from the beginning to the end of a reporting period.

Formula: (1) - (Customers Lost in a Given Period / Number of Customers at the Start of a Period) = (Customer Retention Rate)
Example: (1,050 − 100) ÷ 1,000 = 95%.
What good looks like: there is no universal number — retention varies sharply by industry and contract model, so benchmark against sector peers and track it by customer segment
Weak → strong: "Keep more customers" → "Lift customer retention from 95% to 96% this fiscal year, owned by the Head of Customer Success via an at-risk account program".

Early Repeat Rate

LaggingBetter when higherMeasures the portion of consumers who make a second purchase within a set amount of time after their first purchase.

Formula: (Number of Customers Who Make a Second Purchase Within X Amount of Time)/(Total Customers Who Made a Purchase During this Time) = (Early Repeat Rate)
Example: 180 ÷ 2,000 = 9%.
What good looks like: there is no universal number — it depends heavily on the time window and how often your product is naturally rebought, so fix the window and beat your own baseline
Weak → strong: "Win the second purchase sooner" → "Lift the 60-day early repeat rate from 9% to 12% within two quarters, owned by the lifecycle marketing lead via a post-purchase email series".

This could be a good indicator of how well companies are converting their one-time buyers into more loyal customers.

Life Cycle Distribution Status

LeadingMeasures where customers fall within the various stages of the customer life cycle.

Formula: (Customers in each stage ÷ total customers) × 100.
Example: 300 in onboarding ÷ 2,000 customers = 15%.
What good looks like: there is no universal number — a healthy mix shows customers moving steadily from onboarding to adoption and renewal, with no stage where accounts pile up or stall
Weak → strong: "Move customers through the life cycle" → "Cut the share of customers stuck in onboarding from 15% to 10% by Q3, owned by the onboarding team lead".

With this measurement, companies can evaluate how many customers they have in each particular stage and the rate at which they are moving.

Rate of Adoption

LeadingThe rate (determined by length of time) at which an innovation is adopted by a given population.

Formula: (Users who have adopted ÷ total target users) × 100, tracked over time.
Example: 1,800 ÷ 3,000 = 60%.
What good looks like: there is no universal number — set an adoption target and date at launch, then track how fast the curve climbs against earlier rollouts
Weak → strong: "Get people to use the new feature" → "Reach 75% adoption of the new workflow among licensed users within 90 days of launch, up from 60%, owned by the product manager".

Renewal Rate

LaggingBetter when higherA good indicator of whether clients find a service useful.

Formula: (Clients Who Renew) / (Total Clients Whose Previous Licenses Came to an End) = (Renewal Rate)
Example: 860 ÷ 1,000 = 86%.
What good looks like: as a revenue-based reference point, bootstrapped B2B SaaS companies with $3M–$20M ARR report median gross revenue retention of 91%, with the top 10% at 100%
Weak → strong: "Get more clients to renew" → "Lift renewal rate from 86% to 90% this fiscal year, owned by the renewals manager".

If this rate is low, companies may need to determine why clients are not renewing the service.

Up-Selling Rate

LaggingBetter when higherThe rate at which customers are converted from purchasing a product to purchasing a more expensive version from the same product family, which increases wallet share.

Winback Rate

LaggingBetter when higherTracks the percentage of churned customers who are successfully "won back" into making a purchase during a given period.

Formula: (Churned Clients Who Repurchase) / (Churned Clients) = (Winback Rate)
Example: 45 ÷ 500 = 9%.
What good looks like: there is no universal number — win-back depends on why customers left, so track it by churn reason and against your own baseline
Weak → strong: "Win back lost customers" → "Raise winback rate from 9% to 12% of churned customers within a year, owned by the retention marketing lead".

Referral Conversion Rate

LaggingBetter when higherMeasures the portion of referral invitations sent that are accepted by their recipients.

Formula: (Converted Referrals) / (Total Referral Invitations Sent) = (Referral Conversion Rate)
Example: 90 ÷ 1,500 = 6%.
What good looks like: there is no universal number — it depends heavily on the incentive and how the invitation is delivered, so test offers and track the rate against your own baseline
Weak → strong: "Make referrals work better" → "Raise referral conversion from 6% to 8% within two quarters, owned by the growth marketing manager via a double-sided incentive".

Numbers & Rates KPIs

Average Number of Referrals per User

LeadingBetter when higherA higher number of referrals per user is likely to lead to more sales, increasing the profitability of each customer.

Formula: (Number of Referrals) / (Number of Users) = (Average Number of Referrals per User)
Example: 600 referrals ÷ 2,000 users = 0.3 referrals.
What good looks like: there is no universal number — most users never refer anyone, so grow the average against your own baseline and track how many referrals actually convert
Weak → strong: "Get more referrals" → "Raise referrals per user from 0.3 to 0.4 within two quarters, owned by the growth marketing lead via an in-app referral program".

Bounce Rate

LaggingBetter when lowerMeasures the number of visitors that access a page on a company website and leave before visiting any other pages.

Formula: (Single-page sessions ÷ total sessions) × 100.
Example: 4,500 single-page sessions ÷ 10,000 sessions = 45%.
What good looks like: lower is better — cross-industry benchmark data put average bounce around ~43% on desktop and ~51% on mobile, with e-commerce averaging ~47%
Weak → strong: "Keep visitors on the site" → "Cut bounce rate from 45% to 40% on top landing pages within one quarter, owned by the web manager via faster load times".

Click-Through Rate

LeadingBetter when higherMonitors how many people click on links in an email.

Formula: (Clicks ÷ impressions) × 100.
Example: 2,000 clicks ÷ 100,000 impressions = 2%.
What good looks like: search ads averaged ~6.64% CTR across industries in 2026, while the #1 unpaid search result earns ~27.6%, so compare like with like by channel and placement
Weak → strong: "Make our ads more clickable" → "Raise CTR from 2% to 2.5% on the main campaign by Q3, owned by the paid media manager via ad-copy testing".

This is a good way to gauge the success of an email campaign and the quality of an email’s content.

Show all 28 Numbers & Rates KPIs

Client Summit Attendance

LeadingBetter when higherCounts the number of people who attend a client event.

Formula: (Attendees ÷ attendance target) × 100.
Example: 270 ÷ 300 = 90%.
What good looks like: there is no universal number — set the target from past turnout and your client base, and track which key accounts attend, not just the headcount
Weak → strong: "Fill the summit" → "Reach 100% of the 300-attendee summit target this year, up from 90%, owned by the events manager".

It could be measured as a percentage of a specific attendance target or of the total client base.

Contact Volume by Channel

LaggingKeeps track of the number of support requests by phone and email.

Formula: (Contacts in a channel ÷ total contacts) × 100.
Example: 3,000 email ÷ 10,000 contacts = 30%.
What good looks like: there is no universal number — the right mix follows what customers prefer and what each channel costs, so watch month-to-month shifts and staff each channel to match
Weak → strong: "Move contacts to the right channels" → "Shift email from 30% to 40% of monthly contacts by Q4 while holding satisfaction steady, owned by the support operations manager".

This allows the organization to not only compare which method customers prefer, but also to track the number of support requests month-to-month.

Customer Complaints

LaggingBetter when lowerHelps companies determine whether innovations are effective in improving the customer experience with their product.

Formula: Count of customer complaints.
Example: 45 complaints in a month = 45.
What good looks like: lower is better, but there is no universal number — normalize complaints per 1,000 customers or transactions, then track the trend and fix the top root causes
Weak → strong: "Reduce complaints" → "Cut monthly complaints from 45 to 30 by year-end, owned by the customer service manager via root-cause reviews".

Direct Traffic

LeadingBetter when higherTraffic to a company's website that occurs from visitors typing in the URL directly (i.e. actively seeking it out).

Formula: Sum of sessions arriving with no referring source in a period.
Example: 12,000 sessions with no referrer last month = 12,000 direct visits.
What good looks like: there is no universal number — rising direct visits usually signal stronger brand recall, but analytics tools also file untagged links here, so tag campaigns before trusting the trend
Weak → strong: "Build brand awareness" → "Grow direct visits from 12,000 to 15,000 a month within two quarters, owned by the brand marketing manager".

Indirect Traffic

LeadingMeasures website traffic that stems from indirect sources, such as clickable email campaigns or referral links.

Formula: Sum of sessions arriving from referral links, email campaigns, and other linked sources in a period.
Example: 18,000 referral sessions + 12,000 email sessions = 30,000 visits.
What good looks like: there is no universal number — grow it against your own baseline by source, and favor referral and email visits that convert over raw volume
Weak → strong: "Get more referral traffic" → "Grow indirect visits from 30,000 to 36,000 a month by Q4, owned by the partnerships manager via co-marketing links".

Number of Reads on Company Blog Articles

LeadingHelps companies determine whether visitors are finding their content useful and which content performs better than others.

Formula: Sum of article page views in a period.
Example: 40 articles × 1,500 average views = 60,000 reads.
What good looks like: there is no universal number — grow reads against your own baseline, and compare articles by reads and the actions readers take next to see which topics earn more investment
Weak → strong: "Get more people reading the blog" → "Grow monthly blog reads from 60,000 to 75,000 within two quarters, owned by the content marketing manager via a search-led topic plan".

Number of Support Requests per Product

LaggingAllows a company to determine which products their customers find easier (and harder) to use.

Formula: (Support requests for the product ÷ customers using it) × 100.
Example: 150 ÷ 1,000 × 100 = 15 per 100 customers.
What good looks like: lower usually signals an easier product, but there is no universal number — compare products against each other and track each one after releases
Weak → strong: "Make the product easier to use" → "Cut support requests for the mobile app from 15 to 12 per 100 customers by Q3, owned by the product manager".

Open Rate

LeadingBetter when higherTracking the number of opened and unopened emails allows companies to evaluate whether an email campaign strategy is successful or not.

Formula: (Emails opened ÷ emails delivered) × 100.
Example: 4,000 opens ÷ 20,000 delivered = 20%.
What good looks like: the all-industry average open rate is ~35.63%, but privacy features in some mail apps inflate opens, so treat open rate as a directional signal and lean on clicks
Weak → strong: "Get more people opening our emails" → "Lift open rate from 20% to 26% within two quarters, owned by the email marketing manager via subject-line testing and list cleanup".

Rank on Search Engines

LeadingCan indicate whether a search engine optimization (SEO) process is effective.

Formula: Sum of ranking positions for tracked keywords ÷ number of tracked keywords.
Example: 1,200 ÷ 150 = 8.0.
What good looks like: lower is better — the #1 unpaid result draws ~27.6% of clicks, about 10× the #10 spot, and only 0.63% of searchers click anything on page two
Weak → strong: "Rank higher on Google" → "Improve average position across 150 tracked keywords from 8.0 to 5.0 within two quarters, owned by the SEO lead".

Rate of Referrals

LaggingCan help illustrate customers' level of satisfaction with a product or service.

Formula: (Number of Referrals in Period) / (Units of Time in Period) = (Rate of Referrals)
Example: 240 ÷ 12 months = 20 per month.
What good looks like: there is no universal number — a steady rise against your own baseline signals satisfied customers, and referrals are worth more when you also track how many become customers
Weak → strong: "Get more referrals" → "Grow referrals from 20 to 30 a month by year-end, owned by the customer marketing manager via a referral incentive".

Redemption Rate

LeadingProvides companies with vital consumer behavior information.

Formula: (Reward Points Redeemed) / (Reward Points Offered) = (Redemption Rate)
Example: 450,000 ÷ 1,500,000 = 30%.
What good looks like: there is no universal number — a rising rate usually means members value the rewards, while a very low rate signals a program customers ignore, so track it against your own baseline
Weak → strong: "Get people using their points" → "Raise points redemption from 30% to 38% within one year, owned by the loyalty program manager via simpler redemption options".

Repeat Customer Rate

LaggingBetter when higherIndicates whether a product or service inspires repeat purchases from customers.

Formula: (Customers That Have Purchased More Than Once) / (Unique Customers) = (Repeat Customer Rate)
Example: 440 ÷ 2,000 = 22%.
What good looks like: a 2024 study of 1,500+ online stores found an average repeat purchase rate of 25% (median 22.5%); rates vary widely by product category, so compare against your peers
Weak → strong: "Get customers to come back" → "Raise the repeat customer rate from 22% to 27% within one year, owned by the retention marketing manager via a loyalty program".

Search Volume for Brand

LeadingBetter when higherCan help companies gauge brand awareness.

Formula: Sum of monthly search queries containing the brand term.
Example: 12,000 branded queries last month = 12,000 searches.
What good looks like: there is no universal number — track it month over month against your own baseline and versus competitors' branded searches, as a sign that awareness work is building demand
Weak → strong: "Get more people searching for us" → "Grow branded searches from 12,000 to 14,000 a month by year-end, owned by the brand marketing manager".

Determined by the number of times that consumers search a brand using search engines.

Sessions per Day

LeadingTracks how many times a customer is using a service or product per day.

Formula: Total sessions ÷ (active users × days in period).
Example: 90,000 sessions ÷ (1,000 users × 30 days) = 3.0 sessions.
What good looks like: there is no universal number — the right level depends on how often the product is meant to be used, so grow it against your own baseline and check that more sessions reflect value, not friction
Weak → strong: "Get people using the product more" → "Raise sessions per active user per day from 3.0 to 3.5 by Q4, owned by the product growth lead".

Visit Frequency

LeadingBetter when higherIf a visitor makes frequent trips to the company website or location, it demonstrates interest in the product or service, which can provide a company with more consumer insights.

Formula: Total visits ÷ unique visitors in the period.
Example: 45,000 visits ÷ 15,000 visitors = 3.0 visits.
What good looks like: there is no universal number — it depends on what you sell and how often customers need it, so grow it against your own baseline and pair it with conversion so extra visits mean real engagement
Weak → strong: "Bring visitors back more often" → "Raise visit frequency from 3.0 to 3.5 visits per visitor per month by Q3, owned by the digital engagement lead".

Number of Customer References

LeadingBetter when higherThe number of customers that brands can rely on to refer others to a product or service.

Formula: Count of customers who have agreed to serve as active references.
Example: 40 referenceable customers = 40.
What good looks like: there is no universal number — aim for enough active references to cover each core segment and use case without overusing the same few customers
Weak → strong: "Get more references" → "Grow the active reference pool from 40 to 60 customers by year-end, covering every core segment, owned by the customer marketing manager".

The more customer references, the better.

Number of Customers

LaggingAllows companies to track the size of their customer base over time.

Formula: Customers at start of period + new customers − lost customers.
Example: 2,000 + 300 − 100 = 2,200 customers.
What good looks like: there is no universal number — the healthy signal is steady net growth, with new customers outpacing lost customers period after period
Weak → strong: "Grow our customer base" → "Grow active customers from 2,200 to 2,500 by year-end, owned by the Chief Revenue Officer via new acquisition and churn reduction".

Number of Customers per Employee

LeadingIndicates the workload per employee and how much bandwidth a company has available for each customer.

Formula: (Number of Customers) / (Number of Employees Serving Those Customers) = (Number of Customers per Employee)
Example: 1,200 customers ÷ 8 employees = 150 customers.
What good looks like: there is no universal number — the right load depends on account size and service model, so set it where response times and retention hold steady as the book grows
Weak → strong: "Scale the team efficiently" → "Support 170 customers per employee, up from 150, while holding retention flat this year, owned by the VP of Customer Success".

Number of New Customers

LaggingBetter when higherAllows companies to track the growth rate of their customer base.

Formula: Sum of net-new customer accounts closed in the period.
Example: 28 net-new accounts in the quarter = 28 new logos.
What good looks like: there is no universal number — set it backward from your growth plan and track it alongside average deal size so new logos are the right-fit customers
Weak → strong: "Win more new customers" → "Win 35 net-new accounts per quarter, up from 28, owned by the new-business sales director".

Number of New Marketing Leads

LeadingDetermines how many marketing leads are added in each period, as opposed to the total number of leads.

Formula: (Total Leads) - (Leads at the Beginning of Each Period) = (Number of New Marketing Leads)
Example: 5,600 − 5,000 = 600 new leads.
What good looks like: there is no universal number — set the target backward from the pipeline you need using your own conversion rates, and judge volume alongside lead quality
Weak → strong: "Generate more leads" → "Add 750 new marketing leads a month by Q3, up from 600, owned by the demand-gen manager".

Number of New vs. Repeat Site Visits

LeadingAllows companies to differentiate their website traffic and generate insights on prospective customers.

Formula: ([1] - [Website Visits by New Visitors]) / (Total Website Visits) = (Number of New vs. Repeat Site Visits)
Example: 6,000 ÷ 10,000 = 60%.
What good looks like: there is no single right mix — a high new share shows reach and a high repeat share shows loyalty, so set the target by goal and watch the trend against your own baseline
Weak → strong: "Bring visitors back" → "Raise repeat visits from 40% to 45% of total within two quarters, owned by the lifecycle marketing manager via an email nurture program".

Time per Website Visit

LeadingCan indicate how engaged a visitor was with the website content during their visit.

Formula: Total session duration ÷ number of sessions.
Example: 250,000 minutes ÷ 100,000 sessions = 2.5 minutes.
What good looks like: there is no universal number — longer visits are good on content pages but can signal confusion on checkout or support pages, so judge it by page type against your own baseline
Weak → strong: "Make the site more engaging" → "Lift average visit time on resource pages from 2.5 to 3.0 minutes within one quarter, owned by the content lead".

Brand Attitude Index

LeadingCustomers' attitudes about a brand can include how positive or negative the customer feels about the brand, as well as how strongly they feel about their conviction.

Formula: Sum of respondents' (favorability score × conviction strength) ÷ number of respondents.
Example: 1,800 ÷ 500 = 3.6.
What good looks like: there is no universal number — survey scales differ, so track the index against your own baseline using the same questions each wave, and watch for strongly held negative views
Weak → strong: "Improve how people feel about us" → "Raise the brand attitude index from 3.6 to 4.0 by the next annual survey, owned by the brand director".

Satisfaction KPIs

43. Comparison of Product with Customer Expectations

LaggingBetter when higherThe product or service should meet or exceed expectations in order to retain customers.

Formula: (Customers who say the product met or exceeded expectations ÷ customers surveyed) × 100.
Example: 340 ÷ 400 = 85%.

44. Customer Effort Score

LeadingThe less effort the customer must expend in order to complete their task or goal with your product, the more they will want to use the product.

Formula: Sum of effort ratings ÷ number of responses (typically 1–7).
Example: 1,500 ÷ 1,000 responses = 1.5.
What good looks like: there is no universal number — confirm your scale direction first (some surveys ask how easy, others how much effort), then track the trend against your own baseline
Weak → strong: "Make support easier" → "Improve the customer effort score from 1.5 to 1.3 (less effort) on post-resolution surveys by Q4, owned by the support operations manager".

45. Customer Satisfaction Index

LaggingBetter when higherHelps gauge a company’s success at meeting customers’ needs.

Formula: (Satisfied responses ÷ total responses) × 100.
Example: 880 satisfied ÷ 1,000 = 88%.
What good looks like: 75–85% satisfied is a common 'good' band, but the trend and the reasons behind your detractors matter more than the absolute score
Weak → strong: "Keep customers happy" → "Maintain a 90%+ CSAT on post-resolution surveys, owned by the support manager, reviewed monthly".

Show all 11 Satisfaction KPIs

46. Customer Satisfaction with a Particular Feature

LaggingBetter when higherCan be used to home in on customers' opinions of a specific feature of a product.

Formula: (Users satisfied with the feature ÷ users rating the feature) × 100.
Example: 260 ÷ 400 = 65%.
What good looks like: there is no universal number — compare each feature with your other features and with its own score before and after a release
Weak → strong: "Improve the reporting feature" → "Raise reporting-feature satisfaction from 65% to 75% after the Q3 release, owned by the product manager".

47. Customer Satisfaction With the Buying Process

LaggingBetter when higherIf a buying process is satisfactory for customers, they may be more likely to remain loyal to the product.

Formula: (Satisfied buyers ÷ total buyers surveyed) × 100.
Example: 350 ÷ 400 = 87.5%.
What good looks like: there is no universal number — track it against your own baseline and by sales channel, and fix the steps buyers rate lowest
Weak → strong: "Make buying easier" → "Raise buying-process satisfaction from 87.5% to 92% within two quarters, owned by the sales operations manager".

48. External Benchmark Survey Ratings

LaggingCompares one organization's customer satisfaction with its competitors' customer satisfaction.

Formula: Your satisfaction score − competitor or industry average score.
Example: 81 − 78 = +3 points.
What good looks like: there is no universal number — compare your score with your industry's American Customer Satisfaction Index average (the national score was 77.0 out of 100 in early 2025), and aim to score above it and widen that lead
Weak → strong: "Beat the competition on service" → "Widen our lead over the industry satisfaction average from +3 to +5 points on the next annual benchmark survey, owned by the VP of Customer Experience".

49. Intention to Repurchase

LeadingBetter when higherThis is the equivalent of a likelihood to repurchase, and many times this is self reported in a survey of the customer near the time of the original purchase.

Formula: (Respondents likely or very likely to repurchase ÷ total respondents) × 100.
Example: 720 ÷ 1,000 = 72%.
What good looks like: there is no universal number — stated intent and actual repurchase can diverge, so set the target against your own baseline and check it against real repeat-purchase rates
Weak → strong: "Get customers to come back" → "Raise intention to repurchase from 72% to 78% on post-purchase surveys by Q4, owned by the customer experience manager".

50. Net Promoter Score

LaggingBetter when higherDetermines how likely customers are to recommend a brand to others, generally represented on a 1-10 scale.

Formula: (Number of Promoters) - (Number of Detractors) = (Net Promoter Score)
Example: 60% promoters − 15% detractors = +45.
What good looks like: any positive score means more promoters than detractors; across 150,000+ organizations the 2025 median is +44 and the top quartile is +72 or higher, though norms vary widely by industry
Weak → strong: "Grow loyalty" → "Raise relationship NPS from +45 to +55 by year-end, owned by the VP of Customer Experience".

A score that qualifies promoters (usually 9-10) and detractors (under 6) would need to be determined in order to calculate this metric.

51. Percentage of Customers Who Are "Very" or "Extremely" Satisfied

LaggingBetter when higherDetermining this metric opens up an opportunity for further surveying into what makes these particular happy customers so satisfied.

Formula: (Customers Who Consider Themselves "Very" or "Extremely" Satisfied) / (Total Survey Respondents) = (Percentage of Customers Who Are "Very" or "Extremely" Satisfied)
Example: 620 ÷ 1,000 respondents = 62%.
What good looks like: there is no universal number — this top-two-box share is stricter than overall satisfaction, so track it against your own baseline and by segment, and watch the gap between merely satisfied and top-box customers
Weak → strong: "Delight more customers" → "Lift the very or extremely satisfied share from 62% to 70% on the annual survey, owned by the head of customer experience".

52. Satisfaction With Interaction

LaggingBetter when higherIndicates customers' average ratings of their satisfaction with an individual service interaction.

Formula: (Sum of response rates from customers who rate their satisfaction on a 1-5 scale) / (Total Survey Respondents)

This is normally determined right after the interaction has occurred.

53. Satisfaction With Services Offered

LaggingBetter when higherMay provide insight into which products or services are doing well and whether to expand the product offering.

Formula: (Satisfied respondents ÷ total respondents) × 100, by service.
Example: 410 ÷ 500 = 82%.
What good looks like: there is no universal number — compare scores across your own services and against last year; a consistently lower-rated service is the one to fix or rethink before expanding
Weak → strong: "Make customers happier with our services" → "Lift satisfaction with the implementation service from 82% to 88% by Q4, owned by the services director".

Common questions

Ted Jackson

Co-Founder & Alabama Native

Co-Founder and Managing Partner of ClearPoint. Former VP at Kaplan and Norton's Balanced Scorecard Collaborative.

Ted Jackson is the co-founder of ClearPoint Strategy, a B2B SaaS platform that empowers organizations to execute strategic plans with precision. Before ClearPoint, he was Vice President at the Balanced Scorecard Collaborative — later the Palladium Group — the firm Drs. Robert Kaplan and David Norton founded to put their framework into practice, where he led global sales and support for its strategy management software and wrote three articles for the Balanced Scorecard Report. A Duke and Harvard Business School alumnus, he brings over 30 years' experience in strategy execution, including 15 years implementing the Balanced Scorecard in the field. Ted works closely with customers to ensure the software meets unique challenges, continually refining the platform with his global expertise.