1 of 26

Lenfest Midterm Elections SMS

November 28, 2018

Hero Digital

2 of 26

Analysis Overview

3 of 26

Summary

3

41% of subscribers responded to the texts by either texting to introduce themselves, more information or asking Jonathan a question

Half of the subscribers clicked on an Inquirer article link and were most interested in the ballot question links and guides to questions asked

10% of Inquirer site visitors hit the meter; 3 ended up subscribing

Survey responses were very positive as 77% respondents felt more informed

Usefulness derived from the experiment was correlated with the amount of engagement

4 of 26

Subscriber overview

4

464 subscribers

27 unsubscribes

79% selected a specific region to receive news about

- Of those who selected, 89% chose PA

41% engaged (191 subscribers) by either responding back to the text to see more information or to ask Jonathan a question

- Unable to tell from the data by person if they clicked on any of the links in Groundsource

- Clicks are from Google Analytics

89%

Unsubscribes from Current Sources list with disabled status

5 of 26

Subscribers joined throughout the experiment

5

Volume of texts sent increased week over week

On average 5 additional subscribers joined per text

Weekly dips in texts sent were due to some users not signing up for PA or NJ regions

6 of 26

Users were more likely to respond to get more information

6

29% of subscribers (133) responded to see more information from a subsequent text

21% of subscribers (99) texted Jonathan back at least once with a question or comment

40 subscribers performed both types of engagement

Most subscribers only texted once to see more information

- Most popular texts were users asking to see more stats from Gun Violence 1-2

- Followed by PA Ballot 1-4

Texted Jonathan back defined as any texts to Jonathan (could have been a comment and not a question)

7 of 26

First week saw 3x higher response engagement

7

Engagement in the first week likely was high due to users checking on the experience

- 75% of Week 1 engagement was driven by people responding to see more stats

Week 2’s decrease in response engagement could have been impacted by the fact the week only served a text 1 or 2 for more info to NJ recipients

Most of Week 4 engagements were around ballots, 5 responded about their experience on Election Day

When normalized for the number of CTAs (introduce yourself, text 1 or 2, tell us about your experience), Week 1 still had the highest engagement

Texted Jonathan back defined as any texts to Jonathan (could have been a comment and not a question). Every outgoing text had messaging calling for texts back.

8 of 26

Subtle CTA and Q&A Article texts had highest engagement

8

On average, texts with subtle language (e.g. “text us”) received twice as many responses on average compared to direct CTA texts with “Text 2”

-4.8 vs. 2.8 responses each

Possibly due to the fact that “Text 2” adds an extra step for the user

Could also be that not all responses were actual questions and users felt more comfortable responding with comments to the texts that did not directly ask for questions

Based on text strategy, the Q&A Articles had the highest CTR, twice as high as the other types

- Driven by high CTR (15%) of the Gun Violence week

-Ballot links had the highest CTRs (30%) but fell into the Regular article category, with 10 lower performing links

Texted Jonathan back defined as any texts to Jonathan (could have been a comment and not a question). Every outgoing text had messaging calling for texts back.

9 of 26

On average, subscribers visited the Inquirer site twice

9

Link tracking only available on Inquirer articles

Half of the subscribers clicked through to the Inquirer site (241)
Drove 70 new users to the site

- New user is device and browser specific

Subscribers were slightly more engaged than mobile traffic to the site during this period from other sources

-SMS traffic had a slightly lower bounce rate

- 66% (SMS) vs. 78% (Other sources)

Subscribers accounted for 529 sessions, spending twice as long on site compared to other sources

- 2:29 (SMS) vs. 1:14 (Other sources)

Viewed an average of 1.8 pages/session

- Compared to 1.5 pages/session from other sources

Viewed 4 total pageviews throughout experiment on average

10 of 26

Subscribers viewed other election content

10

Aside from the article they landed on, subscribers were most likely to view other political news articles

Top articles viewed was the Voter Guide and results, expected due to way that suggested/recommended content appears on the page of the articles subscribers landed on

11 of 26

10% of Inquirer visitors hit the meter

11

23 visitors hit the Inquirer meter

The meter was triggered 63 times

Some users hit the meter early on in the experiment

3 visitors subscribed, all who had hit the meter

They all subscribed on November 8, after the Election rather than throughout the experiment

12 of 26

Survey Responses

13 of 26

Respondent overview: Primarily PA Inquirer Subscribers

13

35 total responses

80% of respondents signed up for PA

-89% of all subscribers signed up for PA

69% of respondents subscribe to Inquirer content
31% of respondents subscribe to both print and online content

Respondents shared similar characteristics with the pre-survey respondents

80%

14 of 26

Political Profile: Politically Active

14

All voted

71% have been voting for 20+ years

One first time voter

To be used as segments in following slides but important to consider sample size bias

15 of 26

83% of respondents want to receive future news via SMS

15

Net future subscription rate to SMS news was +77%

- (Yes-No)/Total

Overall, all segments based on voting history length would want to receive news based texts

4-10 and 20+ voting length groups had slightly lower net future subscription rates likely due to having more respondents

- Everyone (3 people) in the never voted and 11-20 year groups answered ‘Yes’

Majority of users were content with frequency of the texts

Net rate for 1-5 scale questions: ((Respondents who rated 4 or 5) – (Respondents who rated 1 or 2)) / Total

16 of 26

77% of respondents felt more informed

16

Majority of respondents rated the question with a 4 or 5

Overall net informed rate +63%

Person who had never voted rated a 3

Of people who felt more informed, 81% of respondents felt that it made voting easier

Of people who do tend to do last minute research, 68% did less

Net rate for 1-5 scale questions: ((Respondents who rated 4 or 5) – (Respondents who rated 1 or 2)) / Total

17 of 26

Users found the option of texting back useful

17

60% of respondents texted Jonathan

Of people who texted Jonathan (20), all except once found the feature at least somewhat useful
Although some respondents didn’t use the feature (12), they still found the option useful to have (8)

Interestingly, 4 respondents didn’t know that they could text Jonathan back

18 of 26

Users also found the links useful

18

83% of respondents at least clicked on one link

Of people who clicked on a link, 93% found them at least somewhat useful

One of the respondents who occasionally clicked on links mentioned that “reading the information on the phone was too tedious”

19 of 26

Usefulness depended on engagement level

19

As expected, users who engaged more found themselves to be most informed as a result of the SMS

Two respondents didn’t engage at all by clicking or texting back

Based on respondents who only either clicked on a URL or texted back, users found the URLs to be more informative

20 of 26

Respondents most interested in breaking news via text

20

Interesting as there are already many new sites that provide notifications on breaking news so we would need to think about how to differentiate

21 of 26

Opinions from long form responses

21

Aside from the survey, a few users also provided feedback when texting Jonathan back
16 survey long form responses

Positive

This was a great service. You earned my subscription. I have enjoyed the coverage, and I subscribed to online plus Sunday. All the best and thanks for all your work to help voters.
I thought you did an amazing job and you covered everything I was concerned about with the candidates. Also an added bonus was I felt like you were actually there and I could just talk to you if I needed too. It was one of the best things I ever signed up for!!!
It was great to receive informative texts and were very easy to navigate for what I wanted to know.

Negative

It felt like information overload, because I also read the daily Philly.com emails and generally live my life on Twitter. But maybe that just means I wasn't the ideal recipient of this campaign.
Reading the information on the phone was too tedious. It’s very hard to read. I would prefer email.
I wish you would have had more information than what I already saw on different news sites. Employment is my number 1 concern.

22 of 26

Recommendations

23 of 26

Recommendations/Considerations

23

Overall positive feedback demonstrates that there is future opportunity with this format

Opportunity exists with sending out breaking news through texts. However, would need to consider how to add differentiated value

Users were most engaged with the guide to submitted questions as it was very unique content. Consider other ways to cover topics and information that is less discussed

A few users felt that there was too much content, future experiments could involve greater visual design testing

24 of 26

Additional Ad-Hoc

25 of 26

Dates of Unsubscribes

25

27 unsubscribes from the 'Sources' data export but without dates

Only 18 that contained the 'stop' text in the raw export with dates

- Groundsource aware of discrepancies

Doesn't look like they were all concentrated around the very beginning of the experiment

Hypothesis is that the 9 missing unsubscribes all came after the survey and simply wasn't tracked in the raw export

26 of 26

Thank You