In modern business, rapid growth is the mantra for success. In a fast-evolving digital market, aggressive user acquisition is the key. And most firms with scalable growth ambitions rely on paid marketing to acquire new users/customers.
Even as monthly marketing budgets exceed six or seven-figures at SMEs, marketers are plagued by a lack of efficiency. Hubspot surveys indicate that at least 40% of digital marketing professionals feel that current paid marketing strategies are not as efficient as they would like them to be.
Spending money on user acquisition, when viewed from the prism of stock market investing, delivers some interesting insights. The role of a modern growth marketer has strong parallels to that of a portfolio manager.
It calls for a fundamental rethink of how marketers approach online campaigns. Instead of a passive, one-dimensional approach to user acquisition, it is time for marketers to take a more active role akin to a stock portfolio manager.
Addressing user acquisition from an investment perspective
For a portfolio manager, the most critical assets are obviously the stocks they manage/own. For a growth marketer, users or customers are the critical assets. And paid campaigns are like a bundle of investments — complete with risk factors, ROI (or rather, ROAS), and a payback/maturity period.
A stock investor has access to stocks, bonds, commodities, ETFs, and other assorted investment vehicles. The modern marketer also has a similar array of diverse options at his/her disposal. Customer acquisition strategies can vary based on channels and target demographics, with finer situational tweaks — think CPC bid strategies, optimization goals, conversion windows.
Rule number one in portfolio management (or even basic stock market investment for that matter) is diversification — never put all your eggs in a single basket. Strategically investing in a wide range of stocks and ETFs brings two benefits — lower exposure to risk and better ROI.
Growth marketers can also adopt similar diversification strategies in paid marketing, since there are high-risk and low risk campaigns to mix and match. For instance, retargeting is a great example of a low-risk campaign. But exclusively focusing on it is sub-optimal, as these campaigns are severely limited by scale.
Targeting high volume, low value customers with low CPA ads is another low-cost strategy that may pay high dividends in the long run — such customers create a viral effect ( K-factor) through both word of mouth and higher App-store ranking driving organic traffic.
In stark contrast, high-value-targets optimization strategies have a higher up-front cost but can pay off early dividends. But the long-term utility of this strategy may be diluted by the trend of hit-and-run users. The solution — spread the ad budget across these different campaigns for a more balanced, and comprehensive risk profile and optimal ROAS.
Understanding the temporal aspect of paid user acquisition
Continuing the parallels between growth marketing and stock investing, it’s time to consider a critical factor briefly alluded to in the previous section — the expected duration of the payback period, and the revenue velocity within the cohort. This is often determined by user behavior and can be briefly summarized into the following broad categories:
- Early birds: these are users who can be easily targetable by the networks as they may usually conduct a purchase event within the conversion window (a matter of days, or even hours). They are definitely purchasers, but they are not necessarily the highest profitability users — there is a chance that at least some of them will turn out to be hit-and-run users.
- Late bloomers: at the opposite end of the spectrum, these are users that demand a fair bit of patience. Revenue can take some time to unveil, but often it is worth the wait as they retain higher levels of interest in the brand and have stronger retention.
This presents a dilemma to marketers — do you swing for the early birds or go long for the late bloomers? Are all early birds created equal, or can you separate them to the hit-and-runs and the actual “Happy Birds” (those that will have strong retention). There is a significant opportunity cost involved in decisions like these. Gunning for the early birds, indiscriminately, using short term events on platforms like Facebook/Google is easy but sub-optimal, as it locks you out of long-term value options.
In contrast, late bloomers may offer better LTV, but represent a higher risk simply due to lack of visibility — how do you identify them efficiently from the crowd is the million-dollar question. Without adequate user segmentation, such campaigns often end up as the marketing equivalent of a Hail Mary — low chance of success due to many users simply not converting at all.
What predictive modeling can bring to paid user acquisition
Platforms like Facebook have evolved significantly from a digital marketer’s perspective. With the addition of free tools like Pixel, custom LTV, and lookalike audiences, businesses now have unprecedented access to user insights on the platform.
Both value optimization and lookalike audiences are touted as the most important factors in user acquisition on modern digital platforms. While that is indeed true to an extent, campaigns based on these tools still suffer from a serious case of diminishing returns.
As the algorithms tend to look at the same group of users, the lookalike audience engine starts reaching its scale limits after a point, with CAC/LTV ratio reaching a breaking point. Marketers are forced to look at alternatives — a different platform with newer audiences, more creative content solutions, and so on.
But another option exists, one that does not involve looking beyond Facebook/Google. The trick is to look beyond the LTV of a monolithic user group. Look more precisely at individual users — like different stocks in the market, users also have different payouts schedules, maturity periods, and so on.
Once you start categorizing users into groups based on the “late bloomer vs early bird” matrix, things start getting a lot more interesting. Now you can have multiple paid acquisition campaigns that target different user groups — each with unique ROAS, LTV, and associated risk factors, much like a stock market portfolio.
This opens up new fields of investment opportunities where using the same tactics of ROI maximization and risk mitigation as a portfolio manager comes in handy. Of course, arriving at an accurate and reliable LTV estimation for individual users is a complicated task.
Like in many other fields, automation can make a world of difference here. Harnessing the power of AI, you can arrive at extremely precise user behavior models, tweaked using business and segment-specific inputs.
Using LTV-based predictive modeling, your business can truly diversify its paid acquisition strategies. It allows for the implementation of stock-market style diversifications strategies in user acquisition, with the potential for better ROAS than ever before.
Diversification in action — a hypothetical user acquisition scenario
Let’s apply the above principle in a basic paid user acquisition scenario. Consider the example of a marketer in charge of an app. It has the widely popular in-app purchase model. The marketer has $2 million at his/her disposal, but only uses $1 million — here’s how you can deploy all $2 million and maintain ROI ( with an assumed average payback period (PBP) of 3 months) :
Consider the following spread, where the marketer only spends $1 million:
a. $50k (CAC $1.5, PBP 1month, ~30K users) on Retargeting churned users
b. $200k (CPI $5, PBP 10months, ~40K users) on Prospecting campaigns optimizing on installs — to fuel future retargeting audiences, and increase app-store ranking and friend invites
c. $750k (CPI $25, PBP 3months, ~30K users) on Prospecting campaigns optimizing on a Purchase event or Value Optimization
Since (a) has scope that has a physical audience limitation, we will focus extensively on (b) and (c)— both, while not constrained by a very limited audience like retargeting is, have limited scope for expansion, as the ROI would become inefficient due to increasing frequency and CAC. The key takeaway — due to saturation, the marketing budget remains underutilized to the tune of $2 million.
With a good user-level predictive LTV model the marketer has the option for a wider, more balanced spread:
a. $50k (CAC $1.5, PBP 1month, ~30K users) — Retargeting on churned users that paid previously
b. $200k (CPI $5, PBP 10months, ~40K users) — Prospecting campaigns optimizing on installs — to fuel future retargeting audiences, and increase app-store ranking and friend invites
c. $1MM (CPI $30, PBP 3months, ~30K users) — Prospecting campaigns running predictive Value Optimization on the predictive LTV of users who convert within 7 days (early birds)
d. $250k (CPI $10, PBP 3.5 months, ~25K users) — Target users that will first convert between 7 days and 30 days
e. $250k (CPI $7, PBP 4.5 months, ~35K users) Target users that will first convert on the 2nd month (late bloomers)
f. $250k (CPI $6, PBP 5 months, ~40K users) Target users that will convert on the 3rd month (late bloomers)
By selectively placing higher bids on long term, high value targets, this spread allows the marketer to increase scale without compromising on ROAS. The result is a portfolio of campaigns that effectively deploys twice the amount as the original, while retaining
a very similar PBP in most campaigns. Even in (e) and (f) where the PBP is higher, it is offset by the increased scale.
With growth, businesses gain access to greater trenches of user data. The no-code, AI-powered Voyantis platform uses predictive modeling to harness this data and provide deeper, more diversified audience targeting. The result is more accurate LTV and user-based models that help keep CAC/CPA low even as your target customer base grows wider.
Such a wider spread of campaign options allows marketers to strategize like portfolio managers. Hedging and diversifying your marketing budget across multiple paid acquisition campaigns helps keep a healthy balance between returns, payback periods and risk at scale. Using Voyantis helps you effectively sidestep the audience scaling limitations of an otherwise excellent marketing platform like Facebook.