The strategic context here is that when Medium membership was introduced, it just changed the ecosystem in different ways. Previously, when you had the app, you could read as much as you want because there was no meter, there was no paywall, it was all just free. Then we introduced Medium membership and all of a sudden, you had to pay beyond a certain limit of stories and we found that was hurting engagement.

We had launched Medium membership in beta form in spring of 2017, but at that point, there was very little content behind the paywall. It didn't really fully pick up until a year after that. At some point in H1 of 2018, the amount of member-only content had just steadily increased through a lot of different initiatives we were doing, so it was a slow ramp up of the intensity of this situation.

The most important thing is for you to be really clear on what the problems are and translate that clarity in your product brief. I think that people can get in their own way by not taking the time to really understand deeply what's going on and push for that high level of clarity on the problems. They start writing and write a lot of fluff and then that stays in place and the project moves forward without the required level of clarity up front.

In other cases, I think PMs can have that clarity in their heads, but then they are either inexperienced at writing in this way or don't take the time to really craft their brief. It actually takes a lot of skill to write a short and clear brief. You don't need to write a novel, but it does take more time to really distill down to clarity. I'll generally write a first draft, I'll leave it, I'll come back to it the next day, and I'll realize, "Oh, I didn't phrase this clearly," or "I said that there were three components, but actually looking at this again, there probably are only two." Pushing yourself to self-edit I think is a really helpful mechanism.

This was all very new territory for us. At Linkedin we had accepted ratios of engagement vs monetization, but at Medium we were figuring it out as we went along. We had to start doing exercises around what you had to believe to see if this [initiative] made sense.

I have a background in BizOps so this type of exercise feels second nature to me. The first thing you need to do is develop some sort of theory of change, and then you create your model.

In terms of building, this project was a bit different because it involved the mobile team and a lot of our development had been on web. It's actually part of why this project took a bit longer–there were the app release cycles, but there were web components to this too. Coordinating much more across web and the app was a new muscle for the team.

We used to have a central mobile team that did all mobile development, but by the time this project happened, we had split them out and mobile devs were allocated to teams by need. We had one Android and one iOS dev assigned to the growth org, so this was essentially our project for them.

Whether to have mobile dev centralized in one team really depends on the org and the people. For us at one point, it did make sense to have all of the mobile engineers together, and then at a certain point, it no longer did.

Caitlin Roman

I used this proposal to get buy-in for a free trial experiment on Medium’s mobile app after the institution of a paywall.

Product proposal - App Free 30 Day (F30) Trial

Users who download the app can only read a limited number of MOC (Member-Only Content) stories before being metered out, and are frustrated that the app is no longer useful to them because so much of the content in the feed is MOC.

 We have seen this sentiment pop up more and more frequently in app store reviews since launching the Quality Pool and FasTrak. Users feel like a bait-and-switch happened, as they downloaded a free app but can’t actually read the majority of the content on it without paying — especially since common hacks like incognito windows or logging out aren’t possible in the app. This problem is particularly acute for new users who aren’t familiar with Medium and our membership model. See appendix for some of these negative reviews.

The number of stories read by app users - who have expressed high interest in using Medium - has trended downward over the last year. 

We believe this is connected to problem #1 - the amount of MOC in the app makes it harder / less appealing for users to read more stories. This is a loss for us in terms of retained users who we could upsell to membership down the road. [Data on stories read per user on the apps over time.]

If we enable new app users* to read more MOC on the app, we will retain more of them as users and can eventually convert them to members at higher rates.


*New app users = users who are downloading the app for the first time. These users may or may not be new to Medium as a whole.

We discussed this initiative in September 2018 and decided not to move forward at that time, largely because we wanted to focus on making the app itself more robust. However, this remains the #1 pain point for new app users, and we believe we can and should solve this problem in parallel with the work that the Mobile teams are doing to improve / reimagine the app itself.

Why should we do this, given that Digest F30 failed?

The Digest version of this initiative launched in Q2 2019 was unsuccessful, but the key difference here is that being metered immediately after downloading the app is a well documented, known pain point - whereas that was not the case with Digest. We are much more optimistic that this app-focused initiative will, in fact, lead to a retention increase.

Most proximate: Increase in stories read per new app user.

Intermediate: More retained app users of the new app user population. (This includes the Activation team’s KR of retained new users, but also includes older users who have downloaded the app for the first time).

Ultimately: More users subscribing to membership from the apps. A long-term increase in new members (even if we lose new members in the short term trial period) as projected from the Member Growth Model.



As a user who signs up or in on app for the first time, I should be able to read all stories on the app without a meter for the next 30 days.

The counter starts as Day 1 when they login on the app for the first time. We do not show the meter on any stories during that time.

Whenever the 30 days is up, we start showing the meter for them, whenever during the calendar month that occurs.

Available to any user who signs up or signs in on the app for the first time for that user account.

As a user in this 30 day period, I should still be metered when I read MOC stories on the Web .

Meter limit still in place for that user on Web.

App reads do not increment the Web meter.

As a user in this 30 day period, I should understand that I am on a 30 day unlimited reading promo when the promo starts.

Show splash screen after login that lets the reader know they have 30 free days of reading

As a user in this 30 day period, I should understand that I am on a 30 day unlimited reading promo during the promotion itself.

Days 1-20: Smaller notification on top of each MOC story

Days 21-30: Bigger UI with Subscribe CTA on top of each MOC story

As a user in this 30 day period, I should be prompted to subscribe when the 30 day period is over.

Day 31+: Onetime splash reminder to subscribe? Or altered meter?

As a user in this 30 day period, I should get push notifications reminding me about my free reading period and encouraging me to read more.

As a person browsing the app store or on Medium mobile web, I should understand that signing up or in on the app gives me 30 free days of stories.

Branch Open in App banners for users without app installed

As a Medium user who does not have the app, I should understand that I can get 30 days of free reading if I download the app.

Promo in metered out email for users without app

 merchandise this more broadly (e.g., in the app store) to start, since this is still a test.

What is the potential impact of this test?

We developed a simple model that assesses the impact on members for each cohort of new app users over the course of their first 5 months since starting to use the app. We make the following assumptions:

Number of new app users in each cohort remains around ~X

We think we’ll observe a long term retention gain from retaining the segment we think 30 days of free reading will resonate with most: those who both churned and were metered out within month 1. This segment is ~X% of the cohort or X metered out new app churners.

Our lift in conversion is determined by the proportion of this segment we think we can successfully change in retention behavior with F30. A X% change in metered out churners’ behavior results in a lift in the cohort retention of ~Xpp.

Outside of this experiment, X new app users usually convert in their first month. In the very worst case no users convert during the trial, they are not retained post-trial, and we maintain flat app user acquisition - we lose all X conversions every month. However we saw from the Digest F30 results [link redacted] that users do convert throughout the experiment, at a depressed rate X% less than control. So the more realistic 

worst case loss launching this experiment at 50/50 for a month is -X conversions (bear case)

 or if we ship, a monthly loss of -2X conversions.

We do not know how users will react but think the mid case, in which the incremental users retain to the same extent and the impact on conversion is essentially neutral, is within the realm of possibility. Based on our assumptions and how the incremental number of retained users convert we create 3 different scenarios to give the range of possible outcomes. The experiment will teach us to what degree behavior changes based on where the results fall within this range.

We then calculate how long it will take after shipping this to 100% that it would take for our long term retention to “pay off” and offset the original conversion loss. Based off of this, in order for it to make sense to ship F30 and have more Members than we would otherwise have had, either:

We retain an incremental users or X% of metered out churners each month, as assumed in the model. But 

we observe a much higher Conversion of the incrementally retained users that is +2x as high as the base case [Bull Case]

, ~X% converted by month 6. This experiment will “pay off” by month 5 and by month 6 we will have cumulatively gained + incremental new members.


We retain ~1.4x more incremental users than the model’s mid case assumption of X%, 

X users or X% of metered out churners each month, from offering free reading. These incremental users convert at the same rate as base conversion, X% converted by month 6. This experiment “pays off” by month 6 and we will have cumulatively gained +X incremental new members.


[Redacted model of Base, Bear, Mid, and Bull Cases]

What is the preliminary estimated level of effort? (from Sept 2018)

Backend: Give first time app users 30 days of free MOC: 

Backfill firstAppInstall timestamp on all existing app users

Mark firstAppInstall timestamp when app config is fetched for the first time

Check firstAppInstall timestamp when determining whether to increment the meter and truncate the post content

: Give first time app users 30 days of free MOC: 

 per platform, though highly dependent on design of in-app product messaging

Don’t show metering UI on post page during this time period

 · To avoid wasting my time and storage space, I researched Medium after I saw the in-app purchases and the ITunes description referenced paid membership being a "perk" and not required to read the articles. After reading two articles, I received a message saying I would have to become a member to read additional articles as my free article limit had been exceeded this month. If those same messages would have appeared in my research instead of after downloading the app, it would have saved me time and now frustration. Maybe I'm clueless or have lived under a rock, but why is there a fee required anyway? It's not a major news publication like their publishing-conglomerate predecessors and they highlight their use of amateur authors which seems to defeat the purpose of a paid subscription? Interesting content, but not enticing enough for a subscription.

 · So almost every story is paid story i.e. you need to upgrade to a paid version, so practically the free app is just a (bad) marketing tactic for paid version and thus useless. I’ve heard a lot about medium but i am deeply dissolved. Great stories but all paid. Deleting this app.

 · This over rated app needs a better way of previewing an article before counting it read. A few overly enthusiastic swipes and other newbie moves cost me a months worth of free articles in my first 3 minutes exploring my new download and I had only just downloaded this app. For real, in less than 3 minutes of opening the app for the first time my first month was gone. Never had such a disappointing app experience. Deleting now.

 · Had some good reads while on this app. Until I opened it up today and found out that you can only read 4 stories per month unless you pay $5 fee to keep reading. Uninstalled, searching for a new FREE reading app.

 · You red a lot the message says. I just had a quick glance into some articles that interested me but I never red more than a view sentences and now you want me to pay. If I knew you wanted to trick me into premium I wouldn’t have downloaded the app.

 · This company is terrible, they lure you in with all these “free” articles then after 10 minutes of reading you start getting blocked by a message “you read a lot, upgrade to continue reading”. This whole site is another “free preview” type scam to take your money. I’m not even sure if the articles on here are valuable, I can’t read any of them! If the site is entirely paid, premium content, which it is, they should be transparent and advertise it as such not mislead and con. Deleting this crap from my phone immediately.

Product Proposal—App Free Trial at Medium

Product proposal - App Free 30 Day (F30) Trial

August 2019

What problems are we solving?

Product hypothesis

Sign up to see more