Website Journey Flow Analysis
Download as PPTX, PDF0 likes50 views
The document provides a summary of website performance and recommendations for improving traffic and conversions for the JAOU website from June to December 2018. Key findings include: - Site traffic has declined since launch but bounce rates are within industry averages. Return visitors spend more time on site than new visitors. - The home page receives the most traffic but visitors spend little time there before leaving. Testing additional landing pages could improve conversions. - Visitors who return multiple times are more likely to convert, suggesting the need for retargeting campaigns. - Top landing pages beyond the home page relate to the enrollment process and attract returning visitors, indicating useful content is driving traffic.
![Top 20 Landing Pages (view 1)
Site Arrival (Landing Page / Non-Bounce) Trending Pattern:
Marketing Campaigns, the main thrust behind creating website traffic,
have been primarily designed to drive traffic to the main home page on
the JAoU website. As a result, the Top 2 primary landing page
destinations for new user visits are as expected and are aligned with the
either the English or Spanish home page.
Other campaign and retargeting activity (national and/or local efforts)
may frequently drive traffic to other landing destinations per defined
objectives, e.g. email retargeting with CTA links to ‘Get Started – Sign
Up’, ‘How to Join’, or cobranded campaign efforts, et al.
One observation for this visitor type we want to pay attention is the
amount of time people are consuming content upon their arrival to the
site and on the landing page before navigating to interior pages as part
of their overall session visit.
For the primary home page, new visitors clearly outnumber return
visitors and are spending a fair, but small amount of time consuming
content before clicking through to another page. This suggests they are
scanning the page for content and information of personal appeal
without diving too deep into the page.
This trend further reinforces our current direction for testing lightly
populated content landing pages.
It’s also interesting to note for non-primary landing pages, where visitors
are spending the most time on page. Getting Started, How to Join and
the FAQ pages are within the Top 5 Landing Pages and are prime
destinations for returning visitors for which we can categorize as visitors
moving through the enrollment journey path.
Other significant journey path pages include Program Overview, Who
Can Join, What You Need to Do, and Benefits of Taking Part. All of
which represent content that visitors are placing a high value.
Entrances represent the number of Visits that started on a specific Web page.
Time Period: Jun 01 thru Dec 31 2018
GA View: 176339261
Ranked by Entrances: Time measurements are in (minutes / seconds) using base 10 decimals.
Deeper analysis is planned in the next review cycle to review influence by device, page load times, session duration, pageviews, page depth, conversion impact, among other key metric attributes to be determined.
Filtered View: This view represents visitors who arrived at the site on a landing page and subsequently navigated into interior pages
for a deeper session and site experience. Approximately 40 percent of all site visits. (bounce visits [single page visits] removed)](https://image.slidesharecdn.com/journey-230830224618-c2f309b7/85/Website-Journey-Flow-Analysis-12-320.jpg)
Website Journey Flow Analysis
- 1. JAOU Website 2018 Performance Review June 01 thru December 31 (v2 Jan 18) Precision Medicine Initiative, PMI, All of Us, the All of Us logo, and “The Future of Health Begins with You” are service marks of the U.S. Department of Health and Human Services.
- 2. Observations & Findings Recommendations ❑ Use 2018 as a benchmark period. Begin applying monthly or quarterly KPI Milestone performance goals to drive overall strategic execution. ❑ Begin setting up campaign tracking for all consortium partners. Develop a reporting mechanism for partners to review their own marketing performance and attribution measured against national program influence. ❑ Place greater marketing campaign emphasis on creating site traffic. More aggressively retarget site visitors for return visits before conversion (expand implementation beyond current email retargeting). ❑ Leverage landing page A/B testing for multiple objectives including lowering bounce rate, with added objectives for improving return visits, content consumption and pageviews. ❑ Adjust the Google default configuration bounce rate. Apply a ‘time out’ formula to better represent a true bounce. ❑ Consider the impact of designing specific culturally correct Spanish landing pages and Spanish content pages for Hispanic heritage self identification. ❑ Reinforce high visitation content pages with greater site design visibility and positioning to aid in the user experience. ❑ Expand the use of co-branded landing pages and consider more Community Partner cobranded landing pages as a means to have specific campaigns drive traffic to specific brand identifiable and trusted brand landing pages. ❑ A large amount of conversions are coming from 12+ site visits which indicates there are users that engage with the site multiple times before first conversion. Consider what landing pages may help a returning visitor learn more about the program and entice them to take action. ❑ Create additional landing pages to vary how users enter the site. As of now most landing page visits are to the homepage which the data suggests might be suboptimal for first-visit conversions. ❑ Implement insights from earlier survey data that may have key insights into landing page creation for optimal demographics targeting, copy creation and imagery. ❑ Expand email list growth strategy past simple newsletter signups. Create a compelling reason for users to engage with the email campaign to drive visitors back to the site. ❑ Consider varying video and thumbnail preview locations throughout each landing page with media content to test for conversion optimization. ❑ Improve communication across creative, analytics and development teams to give a more accurate representation of the entire journey through each channel that drives traffic to JAOU. ❑ The Return Path for visitors is 50% lower when they have had no prior exposure to the program. Further supporting the argument of advertising retargeting or email list growth to drive traffic back to JAOU.
- 3. DRC ENROLLMENT COUNTS Program Conversion Trend Patterns
- 4. Enrollment Count Trendlines Time Period: Jan 01 thru Feb 28 2019 DRC Dataset: as Tracked and Reported by Sage Bionetworks & Scott Sutherland. Blended DRC Dataset + GA View: 176339261 … (GA datapoints are not absolute which do not remove traffic from bots or stakeholder blacklist IP addresses, but can be used with statistical confidence to identify trending patterns.) DRC Enrollment Counts Trending Pattern: Updated enrollment data as reported from the DRC reflects moderate conversion performance with highs and lows per any given month. We can see as the year came to end, conversion counts weakened from a previous (5) five month steady growth pattern. However, moving into 2019, we see a strong reverse trend for enhanced performance. The ‘quality ratio’ (measures account to conversion ratio) has remained reasonably steady month over month and is performing extremely well with an overall 83 percent average. This affirms that when people make the personal commitment to create an account, they are following through in high numbers to complete the Consent process. It also reveals we are securing quality prospects (leads) via the various marketing channels and program events. Furthermore, it’s a good indicator that when candidates arrive at the site and consider enrollment, they are finding sufficient content to help make a personal decision to move trough the enrollment path. The objective moving forward remains to be the need to create a steady stream of new and return website traffic. Our data shows that there is a moderate correlation to site visitors vs. enrollment count correlation (see chart below). 2018 Roundup Counts 177,115 (Accounts / Registrations Created) ꟷ 146,447 (Consent Completions) 83% (Avg. Consent Rate / Quality Ratio)
- 5. Enrollment vs. New User Site Visits Trendlines Time Period: Jun 01 thru Dec 31 2018 DRC Dataset: as Tracked and Reported by Sage Bionetworks & Scott Sutherland. Blended DRC Dataset + GA View: 176339261 DRC Enrollment Counts Trending Pattern: There doesn’t appear to be a strong correlation with new (unique) site visitor site traffic and registrations. Consents closely map to registrations representing that when an enrollee commits to making an account, they move on to the Consent process at an 83 percent quality of lead ratio. Compared to a published report by Smart Insights (19 Aug 2018), industry conversion rates for the Health industry sector was 12.3% for the best conversion rate and 2.8% as the median conversion rate. Our own Conversion Rate of 15.6% (based on unique site visitors) surpasses the closest industry average we have identified to benchmark against and represents the program is doing extremely well for this KPI. https://www.smartinsights.com/ecommerce/ecommerce- analytics/ecommerce-conversion-rates/ Conversion Rate (Accounts Created divided by New (unique) Users. Raw Google Analytics user visit counts does not remove traffic from bots, spiders or blacklisted IP addresses. n = 743,899.
- 6. SITE TRAFFIC Acquisition Trend Patterns
- 7. Baseline KPIs Aggregate Baseline Site Visit Metrics: Site visits continue to show fluctuation month over month with an overall downward trend pattern after the initial launch. Of the core tracking metrics, the objective should be to set monthly or quarterly target goals to ensure consortium marketing efforts are aligned and optimized for reaching month over month site traffic growth milestones. It would also be beneficial to be put into place monthly UTM referral tracking for all consortium partners so that they may review their own marketing attribution performance as to the impact on site traffic. Overall, new users represent the majority of site traffic with return visitors accounting for 22 percent of all monthly website traffic. Sessions, pageviews per user, and average session duration are all respectable with average pageviews per user increasing over time. The average Bounce Rate for the time period represented is also within levels on the low range of similar nonprofit websites (60-75%) - further representing many new visitors are arriving to find information and content expected. Moving into 2019, these early 2018 performance metrics should be used as benchmarks to set track and monitor core KPI monthly milestone targets. Time Period: Jun 01 thru Dec 31 2018 Average Session represented in minutes + fractional minutes. Multiply fractional decimal minutes by (60) to convert to seconds. GA View: 176339261 Raw Google Analytics user visit counts does not remove traffic from bots, spiders or blacklisted IP addresses. n = 743,899.
- 8. Website User Visits Traffic Site Visit Trending Pattern: Residual influence carried over into June from Launch, but has since declined. Ideally, we would like to see a continuing upward trend (even with modest month over month improvement) as program awareness grows among the general population. Yet we are actually seeing a flat monthly trendline pattern for both new and returning users with the exception of the August anomaly. Sustaining month over month growth is critical for both user types. Placing greater emphasis on traffic growth as one measurable campaign KPI objective (for all partners, campaigns and channels) ensures we are continually optimizing content, messages and CTAs for growing program awareness while enriching our understanding for content and message resonance as part of the enrollment journey path. Although not represented here (see next slide series), Bounce Rates are a leading indicator for understanding site traffic and landing page content design and influence, (i.e., are our landing pages delivering on pre-arrival expectations?). Industry nonprofit benchmarks for similar programs reveal we essentially have one opportunity (one site visit) to make an initial impression for a person to decide whether to return and to act on the call to action. Therefore, a goal moving into 2019 should be to leverage the landing page A/B testing to gauge and measure (as one objective) if simplified landing pages further decrease our bounce rates and how they may impact return visits, content consumption and deeper site navigation. Overall, published benchmarks bounce rates for Non-Profits typically range between 60% to 75%. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Raw Google Analytics user visit counts does not remove traffic from bots, spiders or blacklisted IP addresses. n = 743,899.
- 9. Bounce Rate + Avg. Session Duration (All Visits) Bounce Rate + Session Trending Pattern: When reviewing Bounce Rates and Session Duration in general, we must review these metrics with special considerations. What we need to understand is the default configurations Google assigns using single pageview visits for these metrics do not statistically represent the real behavior of our user for a site visit. Therefore, we need to consider what is a bounce? How should we look at session duration? The JAoU web design uses expansive content on the primary landing page, and many first visitors may not navigate beyond the entrance page. For a first time visitor, this is totally understandable. Therefore, what we must determine is how much time in a visit constitutes a bounce and adjust the bounce rate formula accordingly. This can be achieved by using a ‘time out’ function in minutes instead of using the standard one ‘pageview’ formula. This will more accurately represent if site visitors are arriving and finding a site and content as expected which will aid us in our own site design and strategic execution. Over time, we can continue to adjust the bounce rate formula to better reflect a ‘true’ bounce and a ‘true’ session duration via close monitoring of the user journey and conversion tracking. The key benefit will be our ability to better determine if our site design and the marketing messaging designed to drive traffic to the site is aligned to user expectations. (Our current average bounce rate for this time period view and as reported by Google Analytics is within low thresholds for nonprofit industry averages at 55 percent.) Average Duration represented in minutes + fractional minutes. Multiply fractional decimal minutes by (60) to convert to seconds. Time Period: Jun 01 thru Dec 31 2018 Average Bounce Rate is the percentage of single pageview visits to a website. Average Session Duration is the reported average time a user spends on the website (min/sec). GA View: 176339261 Clarification note: Google Analytics bounces are counted as single pageview site visits, and Session Duration time calculations remove bounced users from the session time reported conveying a potentially misleading and inaccurate representation. Avg. Session Duration for this time period view is 7 minutes & 17 seconds (00:07:17). Should visitors spending a fair amount of time on the landing page gathering information be considered a bounce? Raw Google Analytics user visit counts does not remove traffic from bots, spiders or blacklisted IP addresses. n = 743,899.
- 10. Bounce Rate + Avg. Session Duration (New Visits) New Visits Bounce Rate + Session Trending Pattern: Reviewing New Visitor site traffic patterns measured against average monthly Bounce Rates and average Session Duration, we can begin to see the nuances when compared to overall (all user site traffic) averages. Both Bounce Rate and Session Duration do modestly move in opposite (negative) directions as expected from overall averages when new users are isolated and when return user visits are removed. However, in large part, users are still arriving at the site to find a site and content they expected. Furthermore, they are spending a fair amount of time on the site during their first visit (average session duration). This data view begins to help us define a new Bounce Rate Time Out formula. The initial new Bounce Rate formula can be set to an arbitrary time out time calculation based on an average of monthly averages, or even a lower time duration consideration based on what we define as a true bounce (e.g. 90 seconds). It would be further recommended this process be repeated and tested over defined time periods to continuously keep the bounce rate calculation as accurate as possible. And once a new bounce rate formula is implemented, we can expect to see our numbers to perform better while giving us a much truer representation of user site visits both for bounce rate and for average session duration. Average Bounce Rate is the percentage of single pageview visits to a website. Average Session Duration is the reported average time a user spends on the website (min/sec). Average Duration represented in minutes + fractional minutes. Multiply fractional decimal minutes by (60) to convert to seconds. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Raw Google Analytics user visit counts does not remove traffic from bots, spiders or blacklisted IP addresses. Using a Time Out function for reflecting a more accurate Bounce Rate would better represent if a user is not finding a site and content they were expecting upon arrival. Raw Google Analytics user visit counts does not remove traffic from bots, spiders or blacklisted IP addresses. n = 743,899.
- 11. Bounce Rate + Avg. Session Duration (Return Visits) Return Visit Bounce Rate + Session Trending Pattern: The value for reviewing Return Visit traffic patterns is to more accurately understand how much users are engaged upon their second or subsequent visit. We can clearly see bounce rates go down and average session duration goes up significantly over new visit patterns. This represents return visitors are navigating to other site pages and diving into the site content much more frequently and with much more consideration. As a new bounce rate formula is considered, adopted and activated, over time we will continue to have a better picture as to what is occurring to help steer strategic directions. As one objective moving forward, we should set monthly or quarterly milestone goals for reducing bounce rates and increasing average session duration for both new and return visits. This will allow us to continually align and optimize site content / design as well as to strategically plan campaign activity to sustain a continuous upward swing in better performing averages. Average Bounce Rate is the percentage of single pageview visits to a website. Average Session Duration is the reported average time a user spends on the website (min/sec). Average Duration represented in minutes + fractional minutes. Multiply fractional decimal minutes by (60) to convert to seconds. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Raw Google Analytics user visit counts does not remove traffic from bots, spiders or blacklisted IP addresses. n = 743,899.
- 12. Top 20 Landing Pages (view 1) Site Arrival (Landing Page / Non-Bounce) Trending Pattern: Marketing Campaigns, the main thrust behind creating website traffic, have been primarily designed to drive traffic to the main home page on the JAoU website. As a result, the Top 2 primary landing page destinations for new user visits are as expected and are aligned with the either the English or Spanish home page. Other campaign and retargeting activity (national and/or local efforts) may frequently drive traffic to other landing destinations per defined objectives, e.g. email retargeting with CTA links to ‘Get Started – Sign Up’, ‘How to Join’, or cobranded campaign efforts, et al. One observation for this visitor type we want to pay attention is the amount of time people are consuming content upon their arrival to the site and on the landing page before navigating to interior pages as part of their overall session visit. For the primary home page, new visitors clearly outnumber return visitors and are spending a fair, but small amount of time consuming content before clicking through to another page. This suggests they are scanning the page for content and information of personal appeal without diving too deep into the page. This trend further reinforces our current direction for testing lightly populated content landing pages. It’s also interesting to note for non-primary landing pages, where visitors are spending the most time on page. Getting Started, How to Join and the FAQ pages are within the Top 5 Landing Pages and are prime destinations for returning visitors for which we can categorize as visitors moving through the enrollment journey path. Other significant journey path pages include Program Overview, Who Can Join, What You Need to Do, and Benefits of Taking Part. All of which represent content that visitors are placing a high value. Entrances represent the number of Visits that started on a specific Web page. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Ranked by Entrances: Time measurements are in (minutes / seconds) using base 10 decimals. Deeper analysis is planned in the next review cycle to review influence by device, page load times, session duration, pageviews, page depth, conversion impact, among other key metric attributes to be determined. Filtered View: This view represents visitors who arrived at the site on a landing page and subsequently navigated into interior pages for a deeper session and site experience. Approximately 40 percent of all site visits. (bounce visits [single page visits] removed)
- 13. Top 20 Landing Pages (view 2) Site Arrival (Landing Page / Bounce) Trending Pattern: Primary Landing Pages remain the top destination for visitors classified as single pageview Bounces. Since these visits are categorized as Bounces using Google’s default configuration, we are unable to review time on page or session duration to determine deeper engagement behaviors. The key takeaway from this view is the percentage of ‘single pageview’ return visits to specific pages which reveal what content is deemed a key consideration and worthy of a return vist in the enrollment journey path. Another insight we can begin to draw upon by comparing the two views for visitor types, there is a lack of Spanish language pages in the top rankings. The question that needs to be asked, “are Spanish language visitors returning to the site to advance in the journey path?” Only one Spanish language page other than the Spanish home page shows up in either list in the Top 20. Does landing page design (and potentially overall Spanish site design) for Spanish visitors matter for self identification? Would a Spanish language landing page specifically built for the Hispanic community perform better? Would greater emphasis on culturally fitting pages to the Spanish visitor impact return visits, session duration and ultimately conversion? Currently, the English and Spanish landing pages and interior content pages are essentially the same using the same content and visual images. A consideration for 2019 is whether A/B testing is warranted for a Spanish re-design that is specifically representative of the Hispanic community. Entrances represent the number of Visits that started on a specific Web page. %Exit represents the number of visits who left the site on the page. GA View: 176339261 11% of all Landing Page visits are for the Spanish landing page. This may represent a much truer representation for Spanish site usage over the Google Analytics reporting based on user language demographics. Filtered view: visitors who have been classified as a bounce). Time Period: Jun 01 thru Dec 31 2018 Ranked by Entrances v v v
- 14. Top 10 Content (Interior) Pages Top Content Pages Trending Pattern: Removing the primary landing pages we can begin to see which pages and content is being valued most after arriving at the site and upon a return visit. A clear observation is much of this content is largely being consumed by return visitors. This represents a logical and natural user journey conversion path for gathering more information after initial program awareness and first site visit. It further reinforces that it is imperative for us to get first time visitors to return to the website to move candidates through the enrollment journey path. A prime objective should be to reinforce these content pages via prominent site design profiling and positioning within the overall user experience. And beyond site design, these pages with high visitation afford opportunities to disseminate verified valued content and messaging across all active campaigns. Another observation when reviewing the exit rate percentage for each content page is that these individual pages are part of a larger user session. This is verification that visitors are navigating deeper into the site for greater program information as part of their personal enrollment consideration. One last observation is that visitors are spending the most time on the ‘Get Started – Sign Up’ page with an average reported time spent at 6 minutes, 20 seconds. But it’s inexplainable as the limited ‘on-page’ content doesn’t necessarily warrant such time averages. This may be a Google Analytics nuance associated with default configurations, which we will need to investigate further. For the near term, it’s advised that all GA time metrics be used only as a single and not exclusive benchmark to gauge differentials aligned with assorted metric views. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Average Time on Page represented in minutes + fractional minutes. Multiply fractional decimal minutes by (60) to convert to seconds. Ranked by Pageviews Of note, Privacy Safeguards does not show up in this list. It may indicate ‘privacy’ is not as big of a concern for enrollees as we originally believed and as we launched the program.
- 15. CHANNEL & CAMPAIGN Attribution Trend Patterns
- 16. Attribution Tracking Disclaimer The current Google Analytics setup does not allow us to track site traffic and/or conversion attribution with a high degree of confidence. Unresolved technical problems associated with UTM links, the lack of consistent UTM use overall among all partners, and reliability issues surrounding Goal Setup and tracking within Google Analytics limits our ability to analyze traffic and conversion attribution with accuracy. As a result, we can provide only limited preliminary Google Analytic views based on traffic and conversion patterns which we have verified are significantly underreported across all campaign activity. This also limits our ability to track and analyze the user journey accurately via attribution throughout the site. Therefore, the analytic views to follow are limited in scope and should be reviewed with caution without any assurances they accurately reflect the true nature of site performance. We advise that the results presented within this section for attribution should only be used as a representation and are not qualified counts. However, there may be opportunities to use these views as a single benchmark going forward. Steps are being applied to correct this in 2019. It is proposed we set up a task force to address all open areas for improvement with Google Analytics, and to re-review the UTM tagging process and implementation which requires cooperation and collaboration among various stakeholders. We are optimistic we can resolve open issues and have much cleaner and more accurate attribution reporting within the next several months.
- 17. Goal Tracking Attribution (Google Analytics) Goal Setup (GA) Attribution Trending Pattern: Goal Setup and Conversion counts from within Google Analytics is under reporting. Therefore, our review of attribution reporting will not reflect actual confirmed counts. As one example, we can clearly see Display media is being under reported by thousands. A data verification of the DRC data reveals GA Goal conversion counts for accounts and completions are under reported by as much as 20 percent. This is the problem the program continues to face. As such, Google Analytics for Goal Conversions can only be used as one reporting tool to loosely gauge performance across the site. Going forward, we will need to augment the Google reporting where possible with other marketing reporting platforms in use or are coming online, and when marketing pixels have been reinstated which we have a greater degree of confidence. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Accounts Created Channel Attribution
- 18. Traffic Attribution (Default Channels) Default Channel Traffic Attribution Trending Pattern: Another word of caution regarding Google Analytics and Default Channel attribution reporting. There are default configuration nuances associated for how Google Analytics arbitrarily applies channel attribution. As a result, the current UTM tagging and lack of tagging in place across many partner marketing campaigns creates a slightly different picture for results by over reporting and under reporting channel attribution results. To take this into consideration, we conducted a data grooming exercise for the last six months. The net result is that we determined the GA reporting is close enough to use as an approximate, single benchmark with the understanding actual counts may vary by channel with some channel counts impacted more than others. After data grooming, Direct and Display Advertising + Paid Social are the leading channels for site traffic attribution. Of note, Organic Social is surprisingly less representing a much smaller part of overall traffic acquisition. Referral traffic needs to be reviewed with the understanding that it comes from both partner activities largely a result of direct backlinks, and via improper channel assignment for online domains that are likely display advertising clicks which are missing associated UTM parameters. Email, as a channel of site traffic influence, is on the lower end of the acquisition scale. Organic Search is consistent, but minimal indicating mass awareness among the general population has yet to reach critical mass. The Other category, when the data is properly groomed and with proper attribution assigned using source/medium parameters is marginal as expected. DirectTrafficattributionincludes anyuserswhoclickeddirectlytothesiteaswellasallunknownsourcetraffic.The‘Other’distinctionisthatuserssessionsdonotmatchanychanneldescription. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 744,354 Site Visits ------------- 100% Overall Site Traffic Attribution ------------- 66,615 Account Conversions
- 19. Traffic Attribution (National Marketing) Marketing Campaigns Traffic Attribution Trending Pattern: National media is the leading campaign and channel of influence for national marketing campaigns. The media buy has been active since launch and runs across display and includes social media (primarily Facebook and Instagram), with continuing assorted testing across other online channel outlets. Cobranded media is also a part of the national media buy and supports the local HPOs and Consortiums. It runs in tandem with the national campaign to complement overall program awareness and site traffic media trends. A consideration to keep in mind and which could increase these numbers is the national media traffic attribution which may be currently assigned to Referral traffic by Google Analytics. Email retargeting site traffic attribution from newsletter subscriptions is smaller, but plays a vital role for return visits. Social media attribution at the national level is virtually non-existent. However, the national social media program to date has yet to implement a consistent tagging process. The impact is that much of national traffic is currently being assigned to unattributed social traffic as reported by Google Analytics. In 2019, we look to change this practice for better, cleaner reporting and analysis. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 224,641 Site Visits ------------- 30% Overall Site Traffic Attribution ------------- 688 Account Conversions Campaign Channel Traffic Acquisition Acct Conversions National Media 113,446 130 Cobranded Media 67,135 28 Email Retargeting 43,923 528 Organic Social 137 2
- 20. Traffic Attribution (DV Partners) Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 141,420 Site Visits ------------- 19% Overall Site Traffic Attribution ------------- 2,089 Account Conversions DV Partner Traffic Attribution Trending Pattern: Three DV partners are accounting for most of the new site traffic among all DV Partners. WebMD, Blue Cross / Blue Shield and Walgreens all have active monthly campaigns across varying channels. As detailed on the next slide, media and paid social constitute the channels for most influence for site traffic. The San Diego Blood Bank, the MEA Journey Bus Tour and the Scripps managed MediaPlanet campaign account for the remaining low number of visits. Of note, Journey Bus Road Exhibit attribution still needs deeper consideration for how best we track attribution both from on-site kiosks and after the initial in-person engagement. DV Partner Traffic Acquisition Acct Conversions WebMD 58,896 840 BCBSA 40,593 14 Walgreens 35,371 27 BCBSA Regence 2,463 1 SDBB 1,879 479 MEA – Journey Bus 1,209 477 MediaPlanet 1,007 251 CBCC 0
- 21. Traffic Attribution (DV Channels) Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Marketing Campaigns Attribution Traffic Trending Pattern: In aggregate across DV Partners, we can see display media and paid social advertising are the biggest traffic creation channels for site visits. Email to date has been used sparingly, but with the notation the San Diego Blood Bank uses it almost exclusively as their primary communications channel. Other channel traffic including Referral backlink traffic from Partner websites are having small influence in a complementary role. Of note, organic social media is not currently a major focus of any partner. Campaign Channel Traffic Acquisition Acct Conversions Display Advertising 87,770 109 Paid Social Media 38,360 47 Referral Backlinks 7,631 660 Email 7,592 1,271 Organic Social 67 2
- 22. HPO Campaign Traffic Attribution Trending Pattern: Overall in aggregate, HPO Consortiums are having much less influence compared to DV Partners. We can clearly see the standouts, i.e., New England, TACH, Arizona- Banner and Wisconsin. However, what these early results reveal is the need for more awareness campaigns and marketing support driving traffic to the HPO landing pages to maximize site visits and enrollments in each of the HPO Consortium regions. All current traffic attribution is identified as Referral traffic. Plans are underway to begin working with each HPO Consortium for applying tracking codes to their marketing efforts. In the coming months, we should start to be able to track HPO activities and attribution much closer. Traffic Attribution (HPO Consortiums) Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 39,213 Site Visits ------------- 5% Overall Site Traffic Attribution ------------- 7,702 Account Conversions HPO Partner Traffic Acquisition Acct Conversions New England 9,357 490 TACH 5,468 2,848 Arizona-Banner 3,204 1,763 Wisconsin 3,205 717 Pennsylvania 2,631 898 Illinois (IPMC) 1,999 603 California 1,202 96 New York 900 134 Southern 623 89 Southeast 466 60 Jackson-Hinds 78 4 HRHCare 4 0 Cherokee Health 1 0 Attributable Conversion Counts
- 23. Traffic Attribution (Community Partners) Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Community Partner Traffic Attribution Trending Pattern: Influence is minimal on site traffic. The standouts include the San Francisco General Hospital Foundation (SFGHF), the National Alliance for Hispanic Health (NAHH), and the WBRC. Most of the current traffic attribution is identified as Referral traffic, with social media being experimented with at some levels. Other channels show up in the mix, but account for less than one percent of the total traffic attribution. Near-term, we will engage this group our new UTM tracking process as it is readied in early 2019. Another consideration may be more Community Partner landing pages which can be the prime landing destination for all partner marketing campaign activites. Attributable Conversion Counts 32,631 Site Visits ------------- 4% Overall Site Traffic Attribution ------------- 252 Account Conversions
- 24. NIH Support Traffic Attribution Trending Pattern: NIH support is being realized for site traffic and conversions via it’s digital online properties. Although representing a small percentage of overall site traffic, the visits coming from the NIH are consistent month over month. Of note, we can see the Director’s Blog is having minimal impact on return traffic back to the JAOU website. Potentially one of the problems, is that on frequent occasions, marketing campaigns may be sending traffic to the NIH.gov site to view the Director’s Blog videos (and possibly other content). Ideally, all marketing campaign traffic should be directed to the JAOU website. As we can appreciate, we want to keep user clicks to a minimum to get users to the JAOU website. Moving ahead, we should begin scrutinizing much more closely as to where JAOU specific marketing is sending site visitors. Monitoring the NIH backlink traffic flow will help us identify which marketing campaigns are reversing the user path, which we can then address individually with the campaign owners. Traffic Attribution (NIH Support) Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 37,682 Site Visits ------------- 5% Overall Site Traffic Attribution ------------- 5,363 Account Conversions Attributable Conversion Counts
- 25. Traffic Attribution (Unspecified Campaigns) Unspecified Campaign Traffic Attribution Trending Pattern: A fair amount of monthly website traffic and conversions are not attributable due to the lack of proper campaign tagging. Although these monthly totals may not be large, when we factor in other non- attributable Direct and Referral traffic, the representation becomes much bigger. In the coming months, the analytics team will be working more closely with various partners to implement better tracking through the use of a renewed UTM tagging effort. However, of note, Social Media is most impacted here. Our own national social media efforts are likely not being given credit for the influence they are creating. In the same mention as above, we will work with the social team to put into place a tagging process and system to make sure they are given credit for their campaign activities. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Unattributable Conversion Counts 21,838 Site Visits ------------- 3% Overall Site Traffic Attribution ------------- 1,330 Account Conversions
- 26. USER ATTRIBUTES Visitor Characteristics Trend Patterns
- 27. Language Use Traffic Patterns Language Use Trending Pattern: Spanish language use as reported by Google Analytics (and which can be used as one benchmark) is not growing. A consideration is whether there are sufficient and sustaining campaigns designed for the Hispanic audience to grow site traffic and participation. Across all marketing touchpoints, more native language campaigns may need to be executed. And ideally, as a continuous targeted effort without starts and stops for reach high Hispanic population regions across the country. As another consideration, our discussion surrounding a strategic Spanish site redesign direction may enhance the Hispanic user experience complete with language and visual imagery speaking directly to the targeted audience. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Spanish Visits by State Region Google Analytics language visit counts are based on user browser settings and the use of cookies. Inaccuracies do exist. n=758,007. Landing Page analysis suggest 11% of all Landing Page visits are for the Spanish landing page. This may represent a much truer representation for Spanish site usage over the Google Analytics reporting based on user language demographics. English, 95% Spanish , 5%
- 28. Gender + Age Visit Traffic Patterns Gender + Age Visit Identification Trending Pattern: There are inherent limitations for collecting aggregate demographics from within Google Analytics. The data views we can collect represent a very small percentage of overall site traffic and should not be used to make strategic directional decisions. Google heavily samples this data from a very small percentage of site traffic (typically less than 10 percent.) Therefore, the sample size is simply too small. In the coming months as we get reinstate our DMP platform, we will be in a much better position to report on accurate site visitor demographics. Data View Limitations 16,010 Site Visits ------------- 2% Time Period Site Traffic Attribution ------------- 15 States Represented Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Google Analytics heavily sampled user demographic visit counts. n=16,010.
- 29. Geographic Visit Conversion Patterns (Top 25 States) Geographic Trending Pattern: We see an identified trend which may have correlation with HPO regional presences and active campaigns underway in those regions, including national cobranded media. Virginia in isolation, as part of the Capital beltway area, could be influenced by its proximity of program stakeholders. Washington state site traffic is influenced by both BCBSA and Washington University which have active campaigns. When reviewing average site performance metrics across all 50 states and the District of Columbia, the Top 15 states surpass the mean averages. Of note, Arizona leads all states across all benchmark metrics with an excellent 3.1 pageviews per session and with an 13 min / 15 sec average session duration. Although these geographic data views may not provide much value for considering how regional results impact site design, it should create some ideation as to how we might be able to tailor content aligned with specific regional populations. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Ranked by Conversion Rate (Accounts Created divided by New (unique) Users. Raw Google Analytics user visit counts does not remove traffic from bots, spiders or blacklisted IP addresses. n=749,237.
- 30. Geographic Visit Traffic Patterns State Region Site Visit Trending Pattern: From the table presented from previous slide, the Top 15 states ranked by number of Accounts Created conversions are all states where there is an HPO presence. In support of the HPOs, the National Media Campaign has active regional media campaigns running in support of the Consortiums which can be one leading influencer for creating site traffic levels. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Raw Google Analytics user visit counts does not remove traffic from bots, spiders or blacklisted IP addresses. n=749,237. State of Texas Site Visit Traffic (Campaign Review Trendline)
- 31. Geographic Conversion (Account) Patterns State Region Site Visit Trending Pattern: From the table just presented (previous slide), the Top 15 states ranked by number of Accounts Created conversions are all states where there is an HPO presence. In support of the HPOs, the National Media Campaign has active regional media campaigns running in support of the Consortiums which can be a leading influencer for creating the site traffic levels. Time Period: Jun 01 thru Dec 31 2018 GA View: 176339261 Google Analytics Goal Conversion attribution counts. Conversion count: 66,281. n=749,237. State of Texas Conversion Attribution (Campaign Review Trendline)
- 32. CONTENT ENGAGEMENTS Site Visitor Trend Patterns
- 33. Top Bounce Pages Top Bounce Pages: Out of all pages the bounce rates are highest for the main homepage (which also accounts for the majority of traffic to the site, so this is an expected result). Some specific landing pages like the San Francisco General Hospital Foundation landing page has a higher bounce rate. Usually, a high bounce rate is a sign that users because they aren’t finding what they are looking for. We can improve bounce rates by analyzing which pages they occurred on and the traffic sources leading to that page. Considering: Were the steps preceding this page visit consistent with messaging, style and imagery? GA View: 176339261
- 34. Newsletter Subscribe Activity Newsletter Signup Event Flow Mapping: Emails have a sizable impact on the conversion of what appears to be a main part of the DV journey. 449 Account Creations originated from email campaigns. The current email list size is ~166K. 13.3% of users that sign up for the newsletter continue on to create an account. This indicates Newsletter Signup activity is tied to higher site conversion activity. Action Item: Conversions should be sent to the Newsletter Subscription or creation of another offer in exchange for contact information that will lead to future nurturing toward Account Creation. Based off of a Google Analytics sample of 8.98% of total interactions on the site. Some data is unavailable due to constraints between Digital Marketing Protocol’s restrictions against cross-site tracking. Newsletter Signup Event Flow is a representation of events in Google Analytics that contribute to conversion.
- 35. Top 10 Videos Top 10 Videos: Videos are a significant step in the user journey, as seen in the next slides, this has an impact on the User Experience through to conversion. GA View: 176339261
- 36. Video Average View Duration & Traffic Sources Video Average View Duration: Most of the video traffic inside of YouTube’s analytics came from native advertising. The rest were site driven with a small minority of traffic coming through suggested videos or YouTube search. When considering content that falls outside of the consent process there are videos that rose to the top. Other top videos that filter out consent videos include: • All of Us Anthem • What is All of Us? • Why All of Us? Why Now? • Mayo Clinic Biobank • A conversation with Bill Gates • The Dish | Update on All of Us’s Genomics Plan Suggestions: For future content considering how new landing page strategies may be enhanced with video content and utilizing existing content throughout parts of the journey. GA View: 176339261
- 37. Video: Impact on Sign Ups/Events Video Event Flow Mapping: Video appears to be a major step in the journey to Login and Signup events on the JAOU site. From the sample taken below 22.7% of all traffic watched a video on the site that led to Logins and Sign Ups. It’s not clear whether these events occur only as a component of the website design. For example, a video Call To Action can influence the clickthrough rate of a particular campaign because they are either at the top of the landing page or are required to advance forward in the process Ways to Improve: - Consider video button and thumbnail preview location throughout each landing page to test for conversion optimization. Video Event Flow is a representation of events in Google Analytics that contribute to conversion.
- 38. Top Landing Page Social Shares Pages Landing Pages Social Shares: Facebook and Twitter lead the referrals for Social Activity on JAOU. A majority of the pages include the main landing pages, All of Us Journey, Getting Started and direct links to the Prevent Health Disparities video. GA View: 176339261
- 39. Visual Site Flow Visual Site Flow: (Note that banner graphics may vary in size, content and creative. The included graphic is only a sample representation.) The entire DV journey from a visual standpoint is worth considering for future site design and marketing messaging. When a user clicks the ad they are sent to the JAOU site where they visit a series of pages leading to the Account Creation step. The visual flow includes representations of the typical journey onto the site. Discovery Research Customer Engagement
- 40. CONVERSION PATHS User Journey Behavior Trend Patterns
- 41. Customer Journey Customer Journey: In the case of supporting the DV journey we must consider the complex multi-stage customer journey that JAOU prospects and participants flow through to completion. With our current and future analysis we will recommend the strengthening of steps which will contribute to additional retargeting efforts. The visual flow includes representations of the typical journey onto the site. consideration account consent referral awareness PR Radio/TV/Print Online Ads Social Email Website Portal FAQ Newsletter
- 42. Top Landing Page URLs for Conversion Top Landing Pages for Conversion: Most site traffic arrives first on the homepage then is driven through the conversion path to Account Creation. 49.16% of conversions occur this way. 13.19% of site traffic arrives at the Get Started page before moving through to Account Creation. GA View: 176339261
- 43. Program vs. Industry Standard Conversion Rates Program vs. Industry Standard Conversion Rates: Based on our Program Conversion Rate of 11% from 600,000 unique visitors and 66,000 sign up conversions we are above the non-profit Industry Standard of 2%. Source: MarketingSherpa Website Optimization Benchmark Survey 11% Program Conversion Rate ------------- 2% Industry Non-Profit Conversion Rate
- 44. Conversion Journey Paths: Direct Conversion Journey Paths: The report shown is generated from conversion paths, the sequences of interactions (i.e., clicks/referrals from channels) that led up to each conversion and transaction. Direct traffic is normally attributed to a user manually typing the site URL into their browser to arrive at the site. When we factor in the known issues with analytics implementation these visits are from an unknown source. These may be visits that arrived to the site from advertising campaigns or other referral sources. GA View: 176339261
- 45. Conversion Journey Paths: Email Conversion Journey Paths: The report shown is generated from conversion paths, the sequences of interactions (i.e., clicks/referrals from channels) that led up to each conversion and transaction. Email is a growing and viable part of the site growth strategy since it is responsible for leading visitors back to the site. In the cases recorded here the users first signed up for the JAOU newsletter before returning to the site to create an account. GA View: 176339261
- 46. Conversion Journey Paths: Social Conversion Journey Paths: The report shown is generated from conversion paths, the sequences of interactions (i.e., clicks/referrals from channels) that led up to each conversion and transaction. Social network activity in this sample timeframe have led to conversions and in many cases make up the overall journey path that a user goes through. GA View: 176339261
- 47. Conversion Journey Paths: Search Conversion Journey Paths: The report shown is generated from conversion paths, the sequences of interactions (i.e., clicks/referrals from channels) that led up to each conversion and transaction. Organic Search traffic may account for portions of the journey that began through prior education of the program or awareness of JAOU from media. GA View: 176339261
- 48. Account Creation Time Lag Conversions Account Creation Time Lag Conversions: Out of 69,142 Account Creation events 61.16% of those had no prior recorded sessions as determined by Google Analytics. 20,715 Account Creation events occurred 12+ days after the first interaction. That behavior indicates either successful advertising retargeting or effective customer journey path behavior back to the JAOU site for conversions. To measure the ROI of our content marketing or social media marketing campaigns, we’ll have to look at the entire conversion paths with multiple sessions, as opposed to the contribution of a single channel with a single session. Time Lag is the number of days it takes for your visitors to complete the final conversion after their first interaction.
- 49. Consent Complete Time Lag Conversions Consent Complete Time Lag Conversions: Out of 51,989 Account Creation events 54.59% of those had no prior recorded sessions as determined by Google Analytics. 17,989 Account Creation events occurred 12+ days after the first interaction. That behavior indicates either successful advertising retargeting or effective customer journey path behavior back to the JAOU site for conversions. To measure the ROI of our content marketing or social media marketing campaigns, we’ll have to look at the entire conversion paths with multiple sessions, as opposed to the contribution of a single channel with a single session. Time Lag is the number of days it takes for your visitors to complete the final conversion after their first interaction.
- 50. Additional Notes It should be noted that much of the information contained within comes from Google Analytics. The platform may not be as accurate as we would like for tracking users and attribution. However, the data sampling collected and the views built on this data can be used as a valid representation for identifying traffic pattern trends without using the numbers presented as absolute. As a result, there may be missing data points we simply can not collect, or data points that are not as accurate as we would like based on this consideration. Part of the problem is that user data can not currently be exported from Google Analytics with applied filters that have been set up in the platform, and which removes traffic from bots, spiders or blacklisted IP addresses. Therefore, the raw data we are using is all inconclusive including all hits to the website. The other problem remains with lingering technical issues associated with UTM tracking where UTM user link association beyond a visitor landing on the site does not remain linked to the user throughout their sessions. This may prevent us (or creates limited visibility) into tracking aggregate attribution of users through the holistic website experience and ultimately through to goal conversions and post conversion. The Wondros analytics team will continue to work with Vibrent to identify and resolve open Google Analytics issues and to work with more accurate data, but to date there may remain limited abilities to collect comprehensive and accurate website performance data for a more detailed analysis. It should also be noted this is a first preliminary review and more work is forthcoming with plans to dig deeper into the data for a more expansive review in the coming weeks and months.