1. + May 2011, London
2. + Outline ? Introduction: Student Non-profit Mentor Analysis Exchange ? Project Objectives ? Key Steps: Define Goals & KPIs Identify & Prioritise Implement ? Learn to use Filters in GA ? Preliminary Findings ? Next Steps ? Learning & Take-Aways 11ededfdsfsf
3. Student: Minyu Bian @minyubian Non–Profit Rep: Anja Ffrench @anjaffrench Mentor: Phil Pearce @philpearce Minyu is a Digital Planner at Agenda21. She has graduated from the University of Southampton with 1st class Masters degree in Marketing Analytics and has worked with Professor Paul Harrigan for Web Analytics research. Anja is the Marketing & Communications Director at ComputerAid. ComputerAid International is a UK registered charity that aims to reduce poverty by providing refurbished computers to developing countries. Phil is Web Analytics Manager at Jellyfish Online Marketing. He has more than 9 years IT experience and has a solid understanding of using various Analytics packages. Introduction 3
4. Introduction – Analysis Exchange Free Web Analysis For Non-Profits World Wide Non-Profit: Computer Aid • Increase donations • Maximise ROI on paid search ads. • Increase e-mail sign-ups & social followers • Improve lead gathering • Drive site engagement Student: Minyu Mentor: Phil Offer Guidance, Support, & Expertise Offer Support Gain Expertise & Knowledge Further development of Skills 4
5. + Picture of Our First Meet Up
6. + Project Objectives 1. Guide the Student and ComputerAid through the identification and tracking of business goals 2. Diagnose installation problems, set-up goals tracking, and verify the data is accurate 3. Support the organisation obtaining a basic understanding of using GA 4. Allow for the second phase project of data analysis, insight and recommendations 6
7. Computer Aid Define Goals & KPIs Identify & Prioritise Implement Key Steps 7
8. 1. Discuss with Anja about business goals 2. Define KPIs 3. Set up a second GA tracker to enable Administrator access 4. Identify Data collection and data integrity issues Define Goals and KPIs 8
9. 9 Define KPIs
10. Macro-Conversion Goal 1: Corporate Donations of PCs Goal 2: Individual Donations of PCs Goal 3: Beneficiary Applications for PCs Goal 4: Volunteers Applications Goal 5: Fundraisng Event signups Micro-Conversion Goal 1: PDF download of Donate.pdf Goal 2: Outclick:Twitter + Facebook +YouTube Goal 3: Click Like or Retweet 10 Define KPIs
11. Identify 11
12. Identify: Using GA SiteScan & LinkSleuth 12
13. Identify: Using GA Dashboards Using the ‘add to GA dashboard’ button: • 12 account setup errors were identified and saved on the dashboard screen • This allowed us to delete the errors reports from the dashboard once they had been fixed Form URLs did not change after being filled in: www.computeraid.org/enquire.asp 13
14. Identify: Using GA Dashboards 14
15. Site Search was also not enabled. Site Search URL: www.computeraid.org/search-results.asp Identify: Using GA Dashboards 15
16. Identify: Using GA Dashboards 16
17. Identify: Using GA Dashboards 17
18. Identify: Using GA Dashboards 18
19. 19 Can we fix it
20. Implement • Change Form pages’ URLs into “/FormName-Thankyou.asp” • Enable Adwords Cost Data in GA • Enable Site Search tracking • Enable 404 error page tracking • Enable tracking of Newsletter, Mailto, and pdf downloads • Add 20 filters to GA • such as exclude “fb_xd_fragment”URL parameter • Activate Google Webmaster Tool • Create an iGoogle Dashboard and GoogleDoc Repository Implement 20
21. ? Set up filters: 1. Pre-defined 2. Custom ? Filters can only be applied to new data (forward looking) ?Aim at a long-term change 21 Phil Taught Minyu How to Use Filters
22. Data’s integrity : a. Exclude Internal Traffic ? Visitor IP addresses ? Visitor ISP Organisation ? Dynamic IP addresses:(1) tag ?utm_source=exclude_me&utm_nooverride=1; (2) use Campaign Source filter b. Include Computer Aid’s Traffic Note:These three filters ? Hostname were applied to test profiles first c. Include Computer Aid’s CPC Traffic before being rolled out to the main ones. ? Campaign Target URL d. Force Source/Medium/Keyword/Referral/Request URI To Lowercase 22 Learn to use GA Filters
23. Customise GA data: 1. Record Email Traffic (From webmail) 2. Apply Filter to separate Social Network (Update specific websites from ‘Social’ to ‘Referral’ when necessary) 3. Put Navigational Search Into same bucket as bookmarked visitors (Campaign Source = Direct_Search & Campaign Medium=None) 4. Separate Broken pages (Place these into the folder /Error_404/ ) 5. Display Full Referral URL (Clear User-Defined field before sending data to the field.) 6. Track the use of Google Translate 23 Learn to use GA Filters
24. Preliminary Findings 24 Service Provider report can be used to identify corporate donations.
25. Preliminary Findings 25 Beneficiary application is highly influenced by email.
26. Preliminary Findings 26 Cheap laptops & used laptops have generated 25 and 12 conversions respectively.
27. Preliminary Findings 27 We can see that used laptops has resulted in 188% ROI over three times more than cheap laptops (58%). Adwords Cost data import was enabled.
28. Preliminary Findings Site Search was enabled. 28 Search Terms can be used to improve site navigation and content. It helps PPC keywords bidding. In this case, we know people actively search for T&C, which was not easily accessible on the site.
29. Preliminary Findings 404 error pages are now shown in Google Analytics. 1.1% of total traffic is passing through a 404 error page 29
30. Preliminary Findings 30 Generic page titles can be customised on different pages to improve natural search rankings.
31. Next Steps • Establish 1 month data benchmark (now Goal data is being captured) • Enable Phone Call tracking • Add thankyou pages for: Individual donations ,Contact Form, Newsletter signups • Add Message from Subject line to remaining mail to emails • Add AnalyticsProEngine (APE) form tracking and/or ClickTale • Enable GA ecommerce on Wordpay event donations callback page. • Define and enable Custom Variables/Custom Slots • Enable Google Website Optimiser onsite testing • CRM integration (GA cookie imported into CRM) • Phil’s PPC super-tag (useful for CRM import) • Setup Advanced segment • Create GA profiles for SEO + PPC + Social + UK + Non-UK • Review Adwords campaign • SEO quick wins. 31
32. Learning Student: Minyu It was Phil who initially recommended The Analysis Exchange to me and encouraged me to participate to learn hands-on experience of Web Analytics. I was intrigued by the idea of learning Analytics and helping non-profits, so I went to our first meet-up in London, during which I got to know Anja and her organisation, Computer Aid. I realised how my support & learning experience would benefit kids who needed computers in Africa. Instantly, I agreed to start this project. During the whole project, Phil has been incredibly patient and guided me to implement GA step by step. He has not only taught me how to identify problems but also offered me additional resources that he has accumulated through years of learning. From this project, I have learned how to identify errors, customise tracking codes, set up goals, apply filters & regex, and glean insights about site user behaviour. Our organisation rep, Anja, has also been amazingly helpful. Every Tuesday, Phil organised Conference calls for us to update the project progress. Anja always spared her time to attend and assisted us for the site change and contact with Web Developers. I really want to say thank you to them, because this valuable learning experience could not have happened without their help. I also highly recommend other students who are passionate about learning Web Analytics to join the Analysis Exchange for future learning. —————- HIDDEN TEXT ——— During this project, Phil has guided me on implementing GA. He has taught me how to identify problems, and how to solving these. I have learned how to identify errors, customise tracking code, set-up goals, apply filters & use regex 32
33. Learning Non-Profit: Anja • I think WA Exchange is a fantastic idea. For a small charity like Computer Aid International, getting expert advice on how our website analytics could be improved would not have been possible without this programme. • Now the changes have been made to our website, I’m excited about the opportunities for improving Computer Aid’s website performance and hopefully improving the rate of PC donations coming into the organisation! • Phil and Minyu are both obviously experts in the area. I would like to thank them both for all of their help and also for their patience with my ignorance regarding all the technical details which they had to talk me through! Mentor: Phil • I first became aware of Computer Aid after watching a documentary on BBC Click. I was impressed by the simple goal of allowing children and students to learn, by providing the tools to educate within impoverished, under-developed societies, and I wanted to do something to help! • At the same time I had register, but I was dormant within the WA exchange, having not completed any projects. • This was a great opportunity for me to impart my knowledge, give a Student a jumpstart into a successful WA career, and be involved in something I could be morally and educationally, proud of – hence here we are today 🙂 33
34. + 34 Appendices: Example Discussion between Student & Mentor ? Minyu: “1. New vs Returning: 81.06% of visitors are new and their bounce rate is 48.01%. In this case, both site relevancy and stickness have to be improved. ” Phil: “I agree. The website needs to be more engaging for new visitors. Possibly adding an embedded YouTube video to the homepage would help. Once AdWords data is shown, we can check that deep-link landing pages are used (i.e not all ppc traffic is sent to the homepage). SiteOverlay and Navigational summary on the homepage might provide some suggestion for why traffic is bouncing. ClickTale which records mouse movements on a page would also help explain the cause for this. ? Minyu: “2. User Defined: the majority of visits are not set. Does that mean most traffic comes from laptop/PC? (You need to explain this to me.) Is it possible for us to track and get value for the most visits?” Phil: “User-defined is an empty field in GA. It is turned off by default and is populated either via a custom filter, or in this instance by adding code to landing pages e.g… setVar(“Guests”). We need to decide on what 6 values should be stored within setVar and CustomSetVar().” ? Minyu: “3. Map Overlay: Nigeria, Kenya, United Stats, & South Africa are popular traffic locations but all of their bounce rates are very high. Perhaps including national flags on the page is a good idea.” Phil: “The Country flag might be a good a/b test – it might improve the application rate for recipients of the PCs in Africa. Also it will be interesting to see if there is Google translate traffic from these countries, and what the language is changed to.” ? Minyu: “4. Languages: English, French, and Spanish are mentioned on the site at the moment but other languages (e.g. German) are also used and their bounce rates are high. Perhaps we can introduce Google Translate to the site? Phil: “To see visitors who translated the website into German, these would be shown as: /google_translate/de/en/. Yes, it will be interesting to see the volume of traffic that lands on non-English pages which contain ^/.*-(fr|esp).asp in the urls.” ? Minyu: “5. All Traffic Source: Source/Medium is not set.” Phil: “I am not sure which field, you have appended on the pdf. Do you mean ‘Source/Medium+UserDefined’ ?” ? Minyu: “6 & 7. Campaign & KeyWord: Adwords is not linked to the GA account.” Phil: “Yes. This is a serious set-up issue for them, as Bidword and SearchQuery data has been lost/missing.” ? Minyu: “8 – 12. Top Content: error pages, goal setups, regular expression in use & on site search tracking.” Phil: “Yes.”
35. + 35
36. + 36
37. + 37 Appendices
38. + 38
39. + 39
40. + 40
41. + 41