In June 2016, I started working on a tool I called Agamemnon: I wanted to build something that would help user acquisition managers project their DAU levels and cash flows based on their performance marketing campaign data.
Doing this is possible in Excel, of course, but it's difficult to incorporate all of the known / knowable metrics that matter in terms of campaign performance into an Excel model:
- Campaign retention: the retention profile of traffic from various sources, projected out over time. This is needed in predicting eg. how many users from a campaign that generated 1,000 DNU will still be active in 10 days' time, or 30 days' time, or 180 days' time;
- LTV contribution by date: most big, sophisticated companies have a robust LTV estimation model in place to help them predict the LTV by acquisition channel in close to real time. But in putting together a cash flow model, one needs this LTV curve by day to project out cash flow contribution for each day the acquired cohort is alive, eg. how much revenue those 1,000 users will be generating in 10 days' time, etc.
Again, building the above into an Excel model is theoretically possible, but it's impractical: retention and LTV curves need to be calculated for each acquisition source and then applied for each cohort acquired over their lifetimes, and then again for each day the campaign is alive. The scope of such an exercise quickly eclipses what is feasible to do in Excel as the number of campaigns and acquisition sources increases. Such a model really needs to be done programmatically.
For that reason, I created Agamemnon: I wanted to build a tool that would let user acquisition managers quickly put together a cash flow model containing many campaigns of many different quality profiles. A cash flow model like this is valuable for a number of reasons:
- It helps user acquisition teams best allocate their budget by source. Agamemnon breaks down ROAS according to campaign retention and LTV profiles, so user acquisition managers know, based on their campaign performance data, which campaigns are generating the highest levels of return -- and, more importantly, the most absolute profit;
- It helps company executives understand when they can expect profit on a monthly basis. User acquisition teams generally focus their work on unit economics (LTV > CPI), but since LTV is calculated over some time horizon, care needs to be taken to balance growth against the company's financial wherewithal with respect to its ability to finance user acquisition upfront from its balance sheet. User acquisition teams are often intimately familiar with their campaign data, but the acronyms they throw around breezily are sometimes totally alien to the people at their companies controlling the marketing purse strings. Agamemnon helps unite the motivations of the finance team (stay solvent by making more money than is spent) and user acquisition team (buy traffic profitably) so that everyone is pulling in the same direction;
- It helps user acquisition teams scenario plan for app launches. App launches involve a lot of uncertainty: how will retention rates change from soft launch to global launch? How will LTV change? How would different types of featuring impact the near- and mid-term cash flows of the app? Does a chart push make sense? Agamemnon was built to help user acquisition managers put together a number of different scenarios quickly and easily to inform difficult yet significant decisions around launch budget allocation.
Agamemnon entered a closed beta in late 2016 and about 250 people participated in it. The complexity and scope of the project increased dramatically beyond the prototype that I built in the summer of 2016 over the course of about three days: I believe it's now a fully-featured product that can meaningfully make user acquisition teams more efficient. Agamemnon is the tool I used to illustrate the concepts in my Mobile User Acquisition 101 video series.
Agamemnon is freemium: its basic functionality is free to use, but premium features can be unlocked via access passes (1 day, 7 days, and 30 days). If you'd like to use Agamemnon's premium features for free for two weeks, follow the instructions on this page.
I'd like to thank all of Agamemnon's beta testers for the invaluable feedback they provided over the past several months, absent which the product would be far less intuitive.