TL;DR
- Landlord is a multiplayer, location based, tycoon game that used real world data to create engagement for players around the world.
- 25M+ downloads, real time worldwide transactions.
- Issue was the scarcity of properties (and its shares) for new players.
- We’ve added systems that monitor and adjust properties around new players.
- D1 retention +2-3pp, massive amount of positive feedback.
The product
- Game based on real life properties - Google Maps / OpenStreetMap APIs; NASA light pollution API as a factor of property value + check in data from Google/Foursquare.
- 25M+ downloads, studio flagship title.
- Players buy, trade, bid on real properties around the world. Properties are divided into 1000 shares that accumulate to 100%.
- My role: Product Manager/Product Owner, led a team of up to 15 people fully liveopsing the title for 6 years.
The loop, and why it choked
Loop worked. More players → more fierce competition for the same properties → rarity increases the value → owning true Apple HQ becomes a status good → engagement and spend grows. Product was more interesting and bid wars were more engaging the more players were playing.
And that’s why it choked. Real properties = finite supply. We can’t build more real properties and we can’t afford to wait for real life developers ;).
The better the loop worked, the emptier the shelf for new players. Especially in big, popular cities. So eventually we’ve reached a state where there was a high probability that starting in a Capital City was making the start of the game unplayable.
For example Berlin (85% of ALL shares in properties bought).
flowchart TD A[More players acquired] --> B[More properties bought out] B --> C[Scarcity rises] C --> |for newcomers| D[Nothing affordable to buy] C --> |for veterans| G[Higher value, better bid wars, more engagement and spend] D --> E[Early drop-off - growth throttles itself]
How we saw it coming
- We had a system that was calculating the amount of shares available in properties throughout the city, it was a live system recalculating 2-3 times per day.
- We had a possibility to check any given area/city on demand and we did that before every big UA spend in targeted countries.
- I saw a lot of negative reviews from players - “I can’t progress because I don’t have cheap properties around me”, “Can’t start playing, all properties around me are bought up”. Technically it was possible to play using marketplace (real time bidding against other players) but it was extremely hard and inefficient at lower levels.
What we changed
1. Geo-scoped surfacing
We’ve been showing the cheapest properties for new players that recently joined the game.
Why: First session is extremely important and it can’t depend on the marketplace feature where players might not even be able to compete.
2. Expanding search radius
Radius was automatically expanded if the game was not able to meet the criteria for new players (at least X properties that have free shares and are in a given price range).
Why: We had to compromise between the “known” properties that player immediately recognizes as something close to their location and player has a thing to buy. I usually focused on the former because of the way our players engaged with the game, but in this case I’ve made an exception.
3. Reclamation from inactive players
The most lucrative properties were sold to the market through our Marketplace or immediately to the bank if a player was inactive for more than 30 days.
Why: Supply was frozen in the portfolio of inactive players so we had to make sure that it wouldn’t rot there for eternity. This was not a perfect feature and it was done as a last measure.
4. Charles Landlord
We’ve created a landlord persona - Charles. He was teaching new players how to play and allowing them to purchase his own artificial property as a FTUE mechanic that was later repurchased by him and the player earned a profit.
Why: I wanted to add more narrative depth into the game and this idea of a cheeky, old, bald landlord was working quite well in our promos and marketing materials. Idea was to have a proper and 100% bulletproof mechanic for our new players so that even if the scarcity of properties was extreme in a specific location that player would still be able to start playing and earn some progress thanks to Charles.
What moved
- Negative reviews such as “I don’t have anything to buy” practically disappeared - closest proof for what we tried to fix.
- Retention D1 rose 2-3pp.
- Additionally (which wasn’t planned): reclaiming properties from inactive-player mechanic - which was set up to launch first day of each month increased the revenue by 2-3x above the average on those days, without any promotion available.
How I’d measure this today
- What we didn’t do: we measured before/after on cohorts, we didn’t do a controlled AB test.
- Why: We didn’t have tools setup, issue was increasing in time so we had to act quickly. I went with my gut that those mechanics would work. The time pressure on the delivery was extremely high.
- What it does not prove: seasonality, a freshness effect, changes in the traffic mix.
- How I would approach it today: OEC, guardrails, decision criteria set up front, holdout.
What I’d do differently
- The share monitoring system should be an integral feature from the very beginning of the game and not added later (we definitely lost a huge amount of players because of this). All of those 4 systems that we introduced were not added at the same time so it was a long and costly production pipeline.
- The reclaiming shares system was a blunt tool that lacked any finesse. We had a multitude of situations where players were coming back after 3-6 months and they were lost immediately because we sold out their portfolio.
- The easiest solution would be to introduce more shares for properties with a specific printing mechanic that would allow us not to sell out properties from inactive players. I did that in Landlord’s successor -> Landlord GO! however it didn’t work quite as planned, which stopped us from doing something similar in Landlord. Properties had a sentimental value to players, and owning 100% of them was a great achievement for players.