How to Drive Website Traffic: 23 Experiments and What to Measure

Correction and revision — 3 October 2026: “proven strategies” has been replaced with experiments because the article did not provide evidence that each tactic delivers results. Example comments claiming a test occurred have been removed.

Useful traffic is traffic from people who can benefit from the destination. Start with the question your page answers, not a target number of visitors. A sale, enquiry or subscription needs a clear definition and its own measurement.

First diagnose the constraint

Discovery: relevant people do not find the page. Comprehension: they arrive but cannot understand the offer. Decision: they understand but lack the evidence, conditions or confidence to act. Delivery: the offer itself fails to satisfy the need.

More distribution mainly addresses discovery. It can amplify the other problems if the destination is weak.

Search and site structure

1. Answer one specific search question

Make the page useful without forcing the reader through a promotional introduction. Check whether its title matches the actual answer.

2. Connect complementary articles

Link where the next page resolves a genuine follow-up question. Use descriptive anchor text rather than a fixed quota of links.

3. Test a Pinterest route

Match the pin’s promise to its destination. Observe recorded outbound visits and downstream behaviour separately.

4. Contribute to a permitted community

Answer a relevant question under the community’s rules. Include a link only when permitted and needed.

5. Adapt an article for another surface

Keep the useful insight and change the format. Do not multiply identical promotional copies and call that a distribution strategy.

6. Improve a vague title

Specify the subject and decision. Treat any click-through change alongside impression and query changes.

7. Clarify the navigation hub

Group resources by user task. A hub is a route through your work, not an asset you control independently of hosting and platform conditions.

8. Offer a useful subscription resource

Provide a relevant sample or checklist. Make subscription terms and privacy handling clear; access to one asset is not unlimited marketing consent.

Formats and return visits

9. Add contextual internal links

Inspect older pages for relevant new reading paths. Avoid duplicate links that add no navigation value.

10. Compare alternatives fairly

State the profile, criteria, geography and date. A comparison can be useful even when it concludes that different users need different options.

11. Write a documented review

Separate hands-on observations from documentary analysis. Disclose commercial interests before the recommendation.

12. Demonstrate one useful idea in video

Give viewers enough value in the video itself. Explain the additional information the full article provides.

13. Share a professional insight

Choose a topic that fits the professional audience. Report genuine observations and label hypothetical cases.

14. Test an alternative social surface

Check whether relevant discussion actually occurs there. Decentralised technology does not guarantee an audience or a click.

15. Maintain durable explanations

Review evolving terms and links. Evergreen means the core remains useful, not that every fact stays current indefinitely.

16. Invite subscribed readers back

Use an appropriate permission basis, a clear reason to read and an easy exit. Avoid sending to scraped contacts.

17. Adopt a sustainable rhythm

Publish when an answer is ready and useful. More frequent publishing is not itself evidence of higher-quality traffic.

Experiments, measurement and economics

18. Try an authorised offline placement

Use a tested QR destination and permission where required. Record visits from that route rather than assuming nearby attention became traffic.

19. Make the next action explicit

Describe the destination and conditions. Measure the intended action, not every button click as a success.

20. Define source measurements

Keep visits, users, clicks and conversions separate. Record period, attribution method and exclusions.

21. Refresh a page with a real defect

Correct outdated facts, broken links and unclear reasoning. Do not change the date merely to suggest freshness.

22. Build depth around a useful subject

Create distinct answers rather than several pages paraphrasing the same claim. Google’s people-first content guidance emphasises useful content; article volume alone is not an authority guarantee.

23. Inspect the economics after the visit

Estimate useful conversions and their actual value. A commercial click is neither a confirmed referral nor realised revenue.

A simple channel comparison

Hypothetical example: route A sends 100 visits and three qualified enquiries; route B sends 40 visits and four. Their enquiry rates are 3% and 10%, but profitability still depends on effort, cost, enquiry quality and later outcomes.

Keep the denominator and period consistent. If the samples are small, describe the result as preliminary rather than declare a permanent winning channel.

Frequently asked questions

Should every article point to a commercial offer?

No. Link only when it supports the reader’s actual task. A useful answer can end with another explanation or no further action.

Where should a beginner start?

Fix the destination, choose one relevant discovery route and keep a measurement log. Use the creator experiment catalogue or our AI-search guide for a narrower next step.

Methodology: proposed experiments, reviewed 3 October 2026. No traffic increase, conversion lift or causal channel ranking is claimed.

— The Rebel Marketer