A cardmaking tutorial can get hundreds of visits without telling you whether anyone managed to make the card. A custom-card service page might get far fewer visits but answer the questions that lead to an enquiry. Page views alone make those pages look comparable, even though they have different jobs. Start by asking what each page should help a visitor do.
Start with a decision, not a dashboard
Analytics tools offer more numbers than most small creative sites need. Before choosing one to track, write down a decision it could help you make. “Should I add a supply list near the top of my stamping tutorial?” gives you something to investigate. “How can I increase engagement?” doesn't. For the supply-list question, you could look at clicks on the list, visits to related instructions, and comments asking which materials were used.
A useful metric passes three tests:
- It relates to the page's purpose. Newsletter sign-ups may matter on a free template page; completed enquiry forms may matter on a custom-card service page.
- You can define it consistently. “Tutorial completion” needs a clear rule, not a changing hunch based on how long someone stayed.
- You can act on it. When the number changes, you should know which page, visitor path, or tracking setup to inspect before editing anything.
That keeps the dashboard manageable and stops a prominent but irrelevant number from becoming a goal.
Assign each page a job
Tutorials, supply reviews, printable instructions, product pages, and contact forms serve different needs. Giving them all the same success metric hides those differences. Group pages by the help they offer.
| Page type | Primary visitor task | Possible signal to track | What the signal cannot prove |
|---|---|---|---|
| Step-by-step tutorial | Follow a technique or finish a project | Clicks to printable steps or a relevant next tutorial | That the visitor completed the card |
| Supply review | Judge whether a material suits a project | Movement to related examples or comparison sections | That the reader bought or liked the material |
| Free template page | Get a usable template | Successful download events | That the file printed correctly at home |
| Custom-card service page | Decide whether to ask about a commission | Qualified enquiry form submissions | That every visitor needed a custom card |
These are starting points, not targets for every site. A tutorial that answers one precise question may do its job without sending anyone to another page. If a reader learns that a particular ink won't work on glossy cardstock, leaving afterward isn't necessarily a failure.
Choose one primary metric for each important page type, plus one or two supporting signals if they help. For a beginner stencil tutorial, you might track clicks on a downloadable cutting guide, then check visits to the troubleshooting section and recurring questions about adhesive. Don't count every interaction as a success; sometimes people are just browsing.
Separate outcomes from diagnostic clues
Outcome metrics record meaningful actions: a template download, a confirmed newsletter sign-up, or a submitted custom-card enquiry. Diagnostic metrics help explain a change in those actions. Think page visits, traffic source, device category, internal-link clicks, or form errors.
Say downloads fall on a light-up card template page. The count tells you what changed, not why. Did the page get fewer visits? Are more visitors using phones? Does the button still work? Did the layout change? Falling traffic calls for a different investigation than steady traffic and a broken download button.
Look at where visitors came from as well as what they did on the page. A social post about a coloring technique may bring different readers than an email linking returning subscribers to a printable pattern. For a fuller look at channel-level decisions, use [how to tell which channels are working for a cardmaking blog] as a separate exercise. Keep page metrics focused on what happens after visitors arrive.
Define an event before counting it
An analytics event records an interaction. Your tool may collect some events automatically; others may need setting up. Either way, check what the event actually records on your site.
Write a short definition for each event you plan to use:
- Name the action. Use “PDF template download started” rather than “engagement.”
- Specify where it counts. Does it cover any PDF, or just the scored-fold card template on one page?
- Check the trigger. A button click may count even if the file fails to open. A successful form submission is different from a click on “Send.”
- Decide how repeats appear. Someone may download a template twice after losing the file. Total downloads and the number of people who downloaded answer different questions.
- Test it yourself. Try the action on a phone and a computer, then check that the expected event appears without a duplicate.
If your platform can't track an action reliably, don't pretend the count is precise. Note the limitation and consider a simpler signal, such as visits to a distinct thank-you page after a confirmed form submission. Check that too: a bookmarked or publicly accessible thank-you page can inflate the count.
Use rates only when the denominator makes sense
A count tells you how many times something happened. A rate puts that count against a defined opportunity. Twenty template downloads from 200 eligible page visits is a 10% download-per-visit rate. It doesn't mean 10% of readers finished a card, and it may include repeat visitors, depending on how visits are counted.
Rates help when traffic changes, provided the denominator fits the question. Divide downloads by visits to the page offering the file, not by all site visits. Compare enquiry submissions with visits to the service page, not visits to unrelated stamping tutorials. Keep the definition the same across periods before calling a rate higher or lower.
Read engagement signals cautiously
Time on page, scroll depth, and bounce-related measures can point you toward a question, but they can't tell you whether someone was satisfied. A reader might leave a tutorial open while cutting cardstock. Another might find the embossing setting they need in seconds. Depending on the setup, a long visit could include idle time; a short one could be entirely successful.
Scroll depth has the same problem. Reaching the bottom doesn't show that someone understood the instructions, and stopping halfway might mean they've paused to make the card. Pair the number with something you can check. If mobile visitors rarely reach a template button below a long gallery, open the page on a phone. Then test whether placing the button closer to the instructions helps.
Site-search terms, comments, and newsletter replies can fill in what the dashboard misses. If several readers ask whether alcohol markers bleed through a particular cardstock, clarify the materials section. There's no need to turn that into a push for longer time on page.
Compare like with like
A Christmas card tutorial, a year-round ink comparison, and a new pop-up card guide won't follow the same pattern of demand. Their raw visits in a single week say little about quality. When reviewing a metric, note the page type, publication date, season, traffic source, and any substantial edits.
For a seasonal page, compare similar periods across years if you have that history, while allowing for changes in promotion and site content. With a new tutorial, early data is better for spotting practical problems—a file that won't open or a confusing materials list—than for judging the topic. A handful of first visits isn't a verdict.
Small numbers can also make percentages look dramatic. One download from five visits is a 20% rate; one more download makes it 40%. Neither figure calls for a sweeping redesign. Record the count beside the rate and see whether the pattern persists. No single traffic threshold makes every comparison meaningful.
Watch for changes you caused in measurement
A sudden spike might reflect a new tracking event, a duplicated tag, your own testing, or automated traffic rather than reader interest. Note when you change analytics settings, add a button, redesign a page, or move a download. If older data used a different definition, mark that break instead of drawing one continuous trend.
Respect visitors while measuring. Collect only what you need for a clear purpose, follow the consent and privacy requirements that apply to your site, and keep names, email addresses, and message contents out of analytics events. A count of completed enquiries is generally more useful on a dashboard than the contents of private requests.
Build a review routine you can maintain
A monthly check is often more useful than watching every graph daily. For each priority page group, keep a short record of its purpose, primary metric, supporting signals, current definitions, and one decision you're considering. Fix obvious technical failures promptly; give ordinary content changes time to show a pattern.
During the review:
- Check that important events and forms still work.
- Look at counts before interpreting rates.
- Compare similar pages or periods, noting seasonal and promotional differences.
- Read a sample of relevant questions or feedback for context.
- Choose one change you can explain and test, rather than editing several elements at once.
If you're unsure which results deserve attention first, [setting realistic goals for a cardmaking business blog] can help you identify the wider goal. Then use your analytics review to ask whether each page supports it.
For a first pass, choose one tutorial with a downloadable pattern. Make “successful pattern download” its primary metric, check that the event fires once when the file opens, and record eligible page visits beside downloads. Keep reports of printing trouble in the same notes: a download event won't tell you that the template printed at the wrong scale.