Imagine Twitter is developing a new location-based feature that enables Twitter users to find restaurants near them and shows how their friends rated the restaurants. How would you launch this feature?
Cracking the PM Interview by Gayle Laakmann McDowell and Jackie Bavaro recommends a structure that is summarized below.
Ask clarifying questions:
- Who is the user?
- What does the product do for the user and how does it do it?
- What is the goal of the feature?
Think about possible product issues:
- What are the strengths/weaknesses of the product?
- What worries you about the product?
- Would the launch goal be to validate product/market fit?
- To maximize number of users?
- To achieve quick profitability?
- To ensure positive reaction up front?
- What is the test bed market?
- Would you use a controlled invitation?
- Would you rollout a limited version?
- Would you try to make the biggest splash?
- What are the pre-launch tasks?
- What are the during-launch tasks?
- What are the post-launch tasks?
- Who are the users?
- Are you building an MVP or full product?
- What distribution channels will you use?
- How will you roll out the product?
- How will you generate buzz?
- Can you build partnerships?
- Are there any risks for the launch?
INTERVIEWEE: Who is the target user?
INTERVIEWER: Young people, mid-twenties to mid-thirties, that tend to eat out often. We think these users would like a feature that shows them which restaurants around them were recommended and rated by their friends.
INTERVIEWEE: What is the goal of the feature?
INTERVIEWER: To increase the frequency of Twitter usage among young affluent professionals, which is a desirable segment to marketers and advertisers.
INTERVIEWEE: How would the feature work?
INTERVIEWER: The new feature will be a new tab bar button. When tapped, it will show a map of the user’s location along with place markers of the nearest restaurants that the user’s friends have rated. If the user taps on a marker, an info page appears with a description, the number of friends who made recommendations, tweets by the user’s friends about their dining experience, and a star rating. The rating will be from one to five stars based on the average number of stars.
INTERVIEWEE: Is it a feature for mobile or desktop or both?
INTERVIEWER: Just mobile.
INTERVIEWEE: Sounds good. There is one thing that worries me about this feature, which is the narrow focus of only using ratings from a user’s friends. I’m assuming our target segment can afford to eat out frequently and thus this feature will be most successful with affluent young professionals in a cosmopolitan city like San Francisco or New York, for example. But these cities have a very large number of restaurants, and it seems unlikely that a user would have a large enough number of friends that would visit and rate these restaurants to make this feature usable. So, for this feature to have a higher likelihood of success, Twitter should open the making of ratings to everyone on the platform rather than only friends of a user.
INTERVIEWER: I see your point, and I agree. So let’s say we open the entering of the ratings to everyone. Talk now about how you would do the launch?
INTERVIEWEE: Before releasing this feature nationally, we would need to test whether this feature is attractive enough and would move the needle in increasing frequency of use of Twitter by the target segment. I think our target segment resides in large affluent cities, so I would start by releasing the feature only to Twitter users in, for instance, San Francisco. We would need to have a proper metrics measurement system in place, which would take some work, so constraining this feature to one city would make data collection and analysis easier to handle.
How much time is there before launch?
INTERVIEWER: About three months.
INTERVIEWEE: Okay. I am going to brainstorm on a list of tasks that I would recommend doing for pre-launch, during launch and post-launch.
There are three main activities I would focus on before launch:
- Create Tweets Before Launch — Making sure that by launch time there is an existing list of tweets and ratings for a user to view is critical.
- Raise Awareness — Target specific marketing channels to raise awareness and create traction with users.
- Analytics System — Define the right KPIs and have an analytics systems ready for tracking pre- and post-launch metrics.
Create Tweets Before Launch
By the time the launch happens, there should be a list of tweets and ratings of restaurants in the system so a user doesn’t see an empty map and feed when he taps on the new feature.
Here are some ideas on how to do that:
- Partner with Zagat to use their ratings with restaurants, and in exchange offer them free marketing. The free marketing can be in the form of a filter option to see restaurant ratings by Zagat or Twitter users. In the beginning, there will be more Zagat ratings, but with time Twitter ratings will climb.
- Partner with Yelp in the same fashion as Zagat — offer Yelp free marketing by adding a filter option to see ratings by Yelp.
- Create a widget that invites people in the target city to participate in the pre-launch program. The widget would allow participants to create reviews of local restaurants. The reviews would include: restaurant name, location and a rating. After the official program launch, the widget would be uninstalled from the Twitter app.
- Partner with restaurants willing to ask their guests to rate them. This could be a marketing opportunity for the restaurant. For example, if a guest rates the restaurant, the guest would receive a surprise treat from the chef. iPads with the app could be provided for the guests to enter their ratings.
In terms of raising awareness about the new feature, I would put in place the following activities:
- Partner with celebrities and ask them to tweet about their dining experience at restaurants.
- Perhaps create videos with Celebrities where they are seen using the new feature in a fun way. In exchange, Twitter can give these celebrities more visibility on Twitter.
- Target trendy magazines like 7×7 for San Francisco and have chefs or restaurant owners talk about the new Twitter feature in interviews.
- Target tech media like TechCrunch and Mashable to write about the new upcoming feature.
- Target food blogs to write articles about the new Twitter feature.
- For all digital channels, it would be important to have a call-to-action (CTA) in order to measure traction. The CTA could be a landing page that shows a video of a person using the new Twitter feature.
For the analytics system, it is important to think about the right metrics that will help measure the success or failure of the feature.
Pre-launch metrics will tell us how much exposure our campaigns have had before launch. For example:
- Number of CTA clicks per day, week, month until launch to monitor whether the awareness campaign is working.
- Number of magazine and digital media impressions until launch that talk about the new feature.
Metrics to measure the success or failure of the launch:
- How many users access the feature a day, week, month?
- Is the number of users growing?
- Are users using the feature at least once a month? This will tell us how sticky the feature is.
- How many users have provided ratings of their experience per restaurant?
- Fraction of users that started a rating tweet but dropped off? This could be a sign that something is wrong with the UI.
- A frequency distribution of tweets about dining experiences across restaurants. Having no tweets for a restaurant would be a sign of low interest in the feature, for example.
- Average number of tweets that talked about a restaurant.
- A frequency distribution of number of ratings across restaurants. Having no rating for a restaurant would not be good.
- Average number of ratings for restaurants.
- I think we need a signifier that would highlight the new feature on Twitter at launch time. A red callout above the tab bar that says “New” would be one idea.
- Send a notification to the user when a celebrity they follow tweets about a restaurant.
- Invite tech publications to the launch announcement.
Analyze metrics and make a determination to continue or discontinue the new feature. For example:
- Has there been any significant increase in frequency of use of Twitter by our target segment (our overarching goal)? If not, then is it because the feature did not have enough users to move the needle?
- If there wasn’t much change in frequency of use, retire the feature.
- Do a postmortem analysis to get some learnings for the future.
In summary, I was presented with the task of managing the launch of a new location-based feature for Twitter. The new feature would allow users to locate restaurants near them on a map and would highlight ratings and tweets made by Twitter users. I suggested expanding the ability to see restaurants ratings from everyone and not just friends of the user, because the feature would risk the possibility of a restaurant not having any ratings available at launch time. Since this is a new feature targeting a specific demographic of young affluent professionals, I suggested restricting the launch to just one affluent city like San Francisco.
I proposed several tasks for pre-launch, during launch and post launch. Among the several tasks proposed, I think the most important task for pre-launch is to make sure there are tweets and ratings of restaurants for users to see by launch day. At launch time, it is important that a CTA is displayed on Twitter to call attention to the new feature. And for post-launch, I recommend waiting six months or so in order to have enough data to see if the feature made a significance difference in increasing the frequency of use of Twitter by our target segment.