Tom Mansell - BrightonSEO - Winning with an evidence-based strategy

Transcript

1. Winning with an evidence-based strategy Tom Mansell Croud @CroudMarketing Speakerdeck.com/TomMansellCroud
2. “It depends…?”
3. “You would make a great politician!”
4. I would follow the ‘guidelines’ Optimise titles https://www.youtube.com/watch?v=0JD55e5h5JM
5. I would follow the ‘guidelines’ Links matter https://www.youtube.com/watch?v=NCY30WhI2og
6. I would follow the ‘guidelines’ Get your tech right https://www.youtube.com/watch?v=j3KgrbiB1pc
7. But ultimately, I was faking it.
8. And so was Google https://searchengineland.com/google-search-ranking-documents-434141
9. Thankfully this is a thing of the past https://searchengineland.com/google-search-ranking-documents-434141
10. The guidance is better https://static.googleusercontent.com/media/guidelines.raterhub.com/en//searchqualityevaluatorguidelines.pdf
11. https://zyppy.com/seo/quality-rating-scorecard/ The lid is being lifted on how this works
12. The lid is being lifted on how this works https://zyppy.com/seo/quality-rating-scorecard/ https://x.com/alex_buraks/status/1618988134850785280
13. And we have a supportive and progressive community https://www.linkedin.com/posts/markseo_thoughts-on-recent-google-up dates-activity-7181615585745600512-rnbh https://www.linkedin.com/posts/markseo_thoughts-on-recent-google-up dates-activity-7181615585745600512-rnbh https://www.linkedin.com/posts/chris-long-marketing_60-second-seo-w hen-a-page-redirect-multiple-activity-7181711262710411264-Czv6
14. All of this means we have more evidence than ever
15. And our approach to SEO needs to reflect this
16. Tom Mansell Hello Director of Organic Performance
17. Getting buy in and implementing recommendations is difficult
18. It’s hard because we use language like this “We need to optimise 100 page titles” “Our competitors have a higher DR than us” “We’re failing on LCP”
19. We should be using language like this “We will deliver £XYZ in incremental revenue”
20. I believe there are two prerequisites Understanding how to get there Identifying the size of the prize
21. I believe there are two prerequisites Understanding how to get there Identifying the size of the prize
22. How we do this Create performance baseline Calculate incrementality Prioritisation Forward looking performance baseline using a time series model Forward looking incremental performance scenario using forecast position and CvR improvements Weighted priority by category/area using a ‘difficulty / opportunity’ calculation
23. How we do this https://moz.com/blog/seo-forecasting-whiteboard-friday
24. To get something like this £3.1m Organic Search 12 Month (Apr 23 - Mar 24) FY24 Achieved
25. To get something like this £3m FY25 Baseline/BAU Projection -£108k Organic Search 12 Month (Apr 24 - Mar 25) -3% / -£108k YoY £3.1m Organic Search 12 Month (Apr 23 - Mar 24) FY24 Achieved
26. To get something like this £3.6m FY25 Incremental Projection (New Strategy) Organic Search 12 Month (Apr 24 - Mar 25) +16% / + £504k YoY +£612k £3m FY25 Baseline/BAU Projection -£108k Organic Search 12 Month (Apr 24 - Mar 25) -3% / -£108k YoY £3.1m Organic Search 12 Month (Apr 23 - Mar 24) FY24 Achieved
27. I believe there are two prerequisites Understanding how to get there Identifying the size of the prize
28. ✓ Consistent NAP ✓ Accurate business hours ✓ Accurate map pin ✓ Average star rating 4+ stars (best, top searches) ✓ Review frequency (sustained over time) ✓ Review recency (reviewed positively in last month) ✓ Product live inventory integration (see what's in store) + more Get a checklist of things YMYL ✓ Date updated on content ✓ Significant content update since last refresh ✓ Data refreshed and upto date ✓ Author present ✓ Co-authors / ‘reviewed by’ present ✓ Single author location ✓ Optimised author pages + more Ecommerce ✓ Facet pages are indexable (size, colour) ✓ Optimised copy blocks on PLPs ✓ Product depth on indexable PLPs (avoiding thin ranges) ✓ Fallback rules in place for metadata ✓ Pagination present ✓ Use of user generated content (reviews) ✓ Controlled brand text ad visibility + more Local
29. Measure and record results using test and learn logs LAUNCH DATE REVIEW DATE DESCRIPTI ON DATA SOURCE MARKET KPI TEST/CONT ROL RESULTS % LIFT
30. LAUNCH DATE REVIEW DATE DESCRIPTI ON DATA SOURCE MARKET KPI TEST/CONT ROL RESULTS % LIFT 16/11/ 2023 16/02/ 2024 Measure impact of impleme nting PLP copy blocks Google Analytics UK Revenue % share of Organic revenue for test pages +£121,000 +10% Measure and record results using test and learn logs Causal Proof (pre/post)
31. Measure and record results using test and learn logs LAUNCH DATE REVIEW DATE DESCRIPTI ON DATA SOURCE MARKET KPI TEST/CONT ROL RESULTS % LIFT 11/09/ 2023 11/10/ 2024 Excluding brand text ads from Paid Search Google Analytics DE ROI Test - DE Control - AT, CH, NL, IE +1.07 +15% Causal Proof (Geo-split)
32. Taking these results to begin to relatively weight impact PLPs PLP copy blocks (0.7) Optimised H1s (0.3) 0.7 0.3
33. Normalising scoring for a clear comparison KPIS WEIGHTED AGGREGATED SCORE Men’s Hoodies and Sweatshirts Editorial Category Product Experience Technical Authority WEIGHTING 1.82 2.13 1.90 1.40 3.70 3.33 2.2 3.57 4.1 3.9 2.90 3.75 4.00 3.8 20% 30% 20% 15% 10% 5% 3.65 3.21 3.2 2.70 3.85 3.32 3.4 DELTA TO LEADER 1.83 1.97 2 1.5 0.15 0.67
34. We assess all of this at scale Then, we use a proprietary AI framework to build a complete picture of what Google sees. This allows us to build up a large dataset that provides a complete picture of the competitive landscape. Client & competitors data store Important third parties Croud Crawler Croud Crawler
35. We assess all of this at scale Then, we use a proprietary AI framework to build a complete picture of what Google sees. This allows us to build up a large dataset that provides a complete picture of the competitive landscape. Rules engine and data for each criteria, dataset assessed by panel results validated/challenged by panel AI framework and validation SEO.max results
36. SEO.max Powered by Croud Evidence-based organic search strategy built from experimentation and automation at scale.
37. It depends…?
38. 1. Calculate Incrementality Identifying the size of the prize
39. Understanding how to get there 2. Embrace Experimentation
40. WIN Download our guide
41. Tom Mansell Thank You Director of Organic Performance https://www.linkedin.com/in/tommansell/