Marketing teams love dashboards. They love adding new metrics, new charts, new widgets. The average enterprise marketing team tracks somewhere between 30 and 50 metrics across their tools. And yet, when the CMO asks "what is working?", the answer is usually a ten-minute explanation with caveats.
The problem is not a lack of data. It is a lack of focus. When you track everything, you decide nothing. Every metric competes for attention, and the ones that matter get lost in the noise of the ones that don't.
With 73% of marketers reporting that budgets are under more scrutiny than ever (HubSpot, 2025), the ability to point at the right numbers and say "this is what is working, this is what is not, and here is what we should do about it" is no longer a nice skill. It is survival.
Here are 15 KPIs that actually drive decisions, organized into five categories. For each one: what it measures, what "good" looks like, how to calculate it, and when it will mislead you.
Acquisition KPIs
These metrics tell you how effectively you are bringing people to your business and how much it costs.
1. Customer Acquisition Cost (CAC)
What it measures: The total cost of acquiring one new customer. Not a lead, not a visitor. A paying customer.
How to calculate it: Total sales and marketing spend divided by number of new customers acquired in the same period. Include salaries, tools, ad spend, agency fees, and content production costs.
What good looks like: This varies enormously by industry and business model. A B2B SaaS company might have a CAC of $200 to $500. A DTC e-commerce brand selling $30 products needs a CAC under $15 to stay viable. The number itself matters less than the trend and the ratio to LTV (more on that below).
When it misleads: CAC becomes unreliable when you average it across all channels. Your organic search CAC might be $12 while your paid social CAC is $180. The blended number tells you nothing useful. Always segment CAC by channel.
2. Traffic by Channel
What it measures: Where your visitors come from, broken down by source: organic search, paid search, social, email, referral, direct.
What good looks like: Healthy businesses have diversified traffic. If more than 60% of your traffic comes from a single channel, you have a concentration risk. Google algorithm updates, ad platform policy changes, or social media reach declines can cut your traffic overnight.
AI analytics tools for marketing fall into four categories: built-in AI features in platforms you already use (GA4, ad platforms), general-purpose AI applied to marketing data (ChatGPT, Claude), dedicated AI analytics platforms (Amplitude, Mixpanel, Tableau),
ChatGPT data analysis works by uploading CSV or Excel files to the Code Interpreter (Advanced Data Analysis) environment, where ChatGPT writes and executes Python code on your behalf to clean, explore, visualize, and interpret datasets. It handles files up to
There are three ways to connect ChatGPT to Google Analytics: exporting CSV files and uploading them to ChatGPT, using the GA4 API through Code Interpreter, and connecting through an MCP server for real-time access. Each method has different setup requirements,
When it misleads: Traffic volume alone is vanity. 100,000 monthly visitors that convert at 0.5% are worth less than 20,000 visitors that convert at 5%. Always pair traffic by channel with conversion rate by channel.
3. Conversion Rate by Source
What it measures: The percentage of visitors from each channel who complete a desired action: purchase, signup, demo request, whatever your primary conversion event is.
How to calculate it: Conversions from channel divided by sessions from channel, multiplied by 100.
What good looks like: Benchmarks depend on your conversion event. E-commerce purchase rates average 2% to 3%. B2B demo request rates average 1% to 3%. Email consistently converts highest (often 3% to 5%), followed by organic search, with social traffic at the low end.
When it misleads: Conversion rate by source ignores assisted conversions. Social media might have a 0.3% direct conversion rate but influence 40% of your conversions through awareness. Check assisted conversion reports in GA4 alongside direct conversion rates. For a deeper look at how to credit each touchpoint properly, see our marketing attribution guide.
Engagement KPIs
These metrics tell you whether people actually find your content valuable once they arrive.
4. Engagement Rate (GA4)
What it measures: GA4 replaced bounce rate with engagement rate. An engaged session is one that lasts longer than 10 seconds, has a key event, or includes at least two page views.
How to calculate it: Engaged sessions divided by total sessions, multiplied by 100.
What good looks like: 55% to 70% for most websites. Blog content tends to be lower (45% to 55%) because readers often get what they need from a single page. Product pages should be higher (65% to 80%).
When it misleads: A high engagement rate on a checkout page might mean users are confused, not interested. Context matters. Engagement rate on informational pages is positive. Engagement rate on pages designed for quick action (contact forms, pricing pages) is not always a good sign.
5. Pages per Session
What it measures: How many pages a user views in a single session.
What good looks like: 2 to 3 pages per session is typical. E-commerce sites often see higher numbers as users browse products. Blogs are lower because search traffic often lands on a specific post and leaves.
When it misleads: More pages is not always better. If users need to click through five pages to find your pricing, that is a UX problem, not an engagement win. Compare pages per session against conversion rate. If they are both rising, great. If pages per session is rising and conversion is flat, people are looking for something they cannot find.
6. Average Session Duration
What it measures: The average time users spend on your site per session.
What good looks like: 2 to 3 minutes for most sites. Content-heavy sites (media, blogs, educational platforms) should aim higher: 3 to 5 minutes. E-commerce sites are often lower because efficient shopping is fast shopping.
When it misleads: GA4 calculates session duration based on the timestamp of the last engagement event, not the actual time on page. If someone reads your blog post for 8 minutes but never interacts with the page, GA4 may record that as a very short session. Scroll tracking and engagement events improve accuracy.
Revenue KPIs
These are the metrics that connect marketing activity to money. They are what your CFO cares about.
7. Return on Ad Spend (ROAS)
What it measures: Revenue generated per dollar spent on advertising.
How to calculate it: Revenue from ad campaigns divided by cost of ad campaigns.
What good looks like: A ROAS of 4:1 (four dollars of revenue per dollar spent) is a common benchmark for e-commerce. B2B companies often target 5:1 to 10:1 because of longer sales cycles and higher deal values. Below 3:1, you are likely losing money when you factor in COGS, overhead, and other costs.
When it misleads: ROAS only measures revenue against ad spend, not profit. A 5:1 ROAS on a product with 20% margins means you are spending $1 to generate $5 in revenue but only $1 in gross profit. Factor in your margins before celebrating ROAS numbers.
8. Customer Lifetime Value (LTV)
What it measures: The total revenue a customer generates over their entire relationship with your business.
How to calculate it: Average purchase value multiplied by average purchase frequency multiplied by average customer lifespan. For subscription businesses: average monthly revenue per user multiplied by average customer lifespan in months.
What good looks like: The absolute number depends on your business. What matters is the relationship between LTV and CAC (next metric). An increasing LTV over time means your retention and upselling efforts are working.
When it misleads: LTV calculations are projections, not measurements, especially for newer businesses. If you have only been operating for two years, you do not actually know your customer lifespan. You are estimating based on current churn rates, which may change. Be honest about the confidence level of your LTV number.
9. LTV:CAC Ratio
What it measures: The relationship between how much a customer is worth and how much it costs to acquire them.
How to calculate it: LTV divided by CAC.
What good looks like: 3:1 is the benchmark. A customer should be worth at least three times what you paid to acquire them. Below 3:1, you may be spending unsustainably. Above 5:1, you might be under-investing in growth and leaving market share on the table.
When it misleads: The timeframe matters. If your LTV is $3,000 over three years and your CAC is $1,000, the ratio is 3:1. But you are paying $1,000 upfront and waiting three years to recoup it. If your business cannot absorb that cash flow gap, a "healthy" 3:1 ratio can still put you under.
SEO KPIs
These metrics tell you how well you are performing in organic search, which for most businesses is the highest-ROI channel over time.
10. Organic Traffic
What it measures: Visitors arriving from unpaid search engine results.
What good looks like: Growth trajectory matters more than absolute numbers. 10% to 20% year-over-year growth in organic traffic is strong for established sites. New sites can see much faster growth in early months as they build authority.
When it misleads: A Google algorithm update can spike your traffic temporarily with no action on your part, or drop it through no fault of your own. Always correlate organic traffic changes with algorithm update timelines. Also, rising traffic from irrelevant keywords (your SaaS blog ranking for a recipe query, for example) inflates the number without business value.
11. Keyword Rankings
What it measures: Your position in search results for target keywords.
What good looks like: Page one (positions 1 through 10) for your most important commercial keywords. Position 1 gets roughly 27% of clicks, position 2 gets 15%, position 3 gets 11%. Beyond position 10, click-through rates drop below 1%.
When it misleads: Rankings are personalized, localized, and volatile. The position you see when checking manually is not what everyone sees. Use rank tracking tools to get aggregated, depersonalized data. Also, ranking for a keyword that nobody searches is meaningless. Always pair keyword rankings with search volume.
12. Domain Authority / Domain Rating
What it measures: A third-party estimate of your site's overall ability to rank, based primarily on your backlink profile. Moz uses Domain Authority, Ahrefs uses Domain Rating.
What good looks like: These scores run from 0 to 100 on a logarithmic scale. Going from 20 to 30 is much easier than going from 60 to 70. A DA/DR above 40 is respectable for small to mid-size businesses. Above 60 is competitive.
When it misleads: Domain Authority is not a Google metric. Google has said they do not use it. It is a proxy for link strength, and while it correlates with rankings, the correlation is not perfect. A site with DA 35 can outrank a site with DA 65 if it has better content, stronger topical authority, and more relevant links for that specific topic.
Content KPIs
These metrics tell you whether your content strategy is working at the individual asset level.
13. Traffic per Post
What it measures: Monthly organic traffic to each piece of content.
What good looks like: Most blog content follows a power law: 10% to 20% of posts drive 80% of the traffic. A "good" post is one that consistently brings 500 or more organic visits per month after the first three months. Top performers bring 5,000 or more.
When it misleads: Traffic per post does not tell you about quality of traffic. A post getting 10,000 visits per month from informational queries with zero purchase intent may be worth less than a post getting 200 visits from high-intent commercial queries.
14. Time on Page
What it measures: How long users spend reading a specific piece of content.
What good looks like: For long-form content (1,500+ words), 3 to 5 minutes indicates real engagement. For shorter content, 1 to 2 minutes. If average time on page is under 30 seconds on a 2,000-word article, people are not reading it.
When it misleads: Time on page suffers from the same measurement limitation as session duration in GA4. If a user reads your full article and then closes the tab, GA4 may not capture the actual reading time unless you have scroll depth or engagement events configured. Always implement scroll tracking.
15. Social Shares
What it measures: How often your content is shared on social media platforms.
What good looks like: Social shares are more useful as a relative metric than an absolute one. Compare shares between your own posts to identify what resonates. Industry benchmarks are unreliable because sharing behavior varies enormously by audience, platform, and topic.
For email-specific benchmarks including open rates, click rates, and revenue per send by industry, see our email marketing benchmarks breakdown.
When it misleads: Social shares do not correlate strongly with revenue or even traffic. A post that goes viral on LinkedIn might get 500 shares and send 3,000 visitors, none of whom buy anything. A technical comparison post with 5 shares might drive $50,000 in pipeline. Shares measure resonance, not commercial value.
The Real Problem: Measuring Everything, Deciding Nothing
The trap most marketing teams fall into is not a lack of metrics. It is metric overload. When every dashboard has 20 widgets and every report has 15 charts, the data becomes wallpaper. People glance at it, nod, and go back to doing what they were already doing.
Marketing budgets sit at 7.7% of company revenue (Gartner, 2025), and 59% of CMOs say that is not enough (Gartner, 2025). When resources are tight, you cannot afford to spread your attention across 50 metrics. You need to pick the ones that drive action and ignore the rest.
Here is a practical approach:
Pick 5 primary KPIs. One from each category above. These go on your executive dashboard and get reviewed weekly. Every other metric is supporting detail that you check when something in your primary five changes.
Set thresholds, not just targets. A target is "increase organic traffic by 20%." A threshold is "if organic traffic drops below 10,000 monthly visits, investigate immediately." Thresholds trigger action. Targets trigger annual reviews.
Connect marketing KPIs to AI analysis. KPIs only matter if they drive decisions, which is the core principle behind data-driven marketing. Data analysis is the second most common use case for AI in marketing, at 30% (HubSpot, 2025). Tools like Ooty Analytics let you connect your marketing data to AI assistants like ChatGPT, Gemini, and Claude so you can ask questions about your KPIs in natural language instead of clicking through dashboards.
Review and prune quarterly. Every quarter, ask: did this metric change any decision in the last 90 days? If not, remove it from your dashboard. A metric that has never prompted an action is decoration, not intelligence.
Building Your KPI Framework
Start here:
Define your business objective. Revenue growth? Market share? Profitability? Your KPIs should ladder up to one primary objective.
Select 3 to 5 primary KPIs. These are the numbers you look at weekly.
Select 5 to 10 secondary KPIs. These are the numbers you check when a primary KPI moves unexpectedly.
Set thresholds for each. What number triggers investigation? What number triggers celebration?
Assign owners. Every KPI has one person responsible for it. Not a team. One person.
Build a single dashboard. Not five dashboards. One. With your primary KPIs visible without scrolling. Read our dashboard guide for how to build one that people actually use. For SEO-specific metrics and stakeholder communication, our SEO reporting guide covers what to include and how often to report.
The goal is not to measure more. It is to measure what matters, act on what you learn, and ignore the rest.
