Save $30 with code THANKYOU.  Click here

The Business Math Behind Instagram’s Algorithmic Feed

Let’s look at the business equation underlying Instagram’s upcoming launch of an algorithmic feed and how it could impact their user metrics and advertising business.

Yesterday I wrote a deep-dive into Instagram and Snapchat by looking at how an algorithmic feed is an important part of building an advertising business, and how its introduction impacts the existing trade between brands and influencers. 1

Let’s look at some of the business math behind Instagram’s product changes.

* * *

The rationale to launch the algorithmic feed was often justified by a desire to show people photos they want but miss. The impact, from a business perspective, would be to:

  • Increase user engagement and increase viewing of posts, from the rationale that a product that delivers more of the images that we want will make use want to view even more images;
  • Increase user retention from a new user perspective, in that an algorithmic feed can make it easier to deliver a quality experience to people faster, rather than forcing them to find the right people to follow from the start, solving the “empty stream” problem;
  • Increase user retention from an existing user perspective, by delivering a better product that keeps people using it longer (although that hardly seems like Instagram’s problem right now);

But contained within those changes is a deeper strategic rationale:

  • Reduce organic reach of brand Instagram accounts, and thus increase their need to buy ads to reach their followers
  • Create new content and / or ad units powered by an algorithmic feed

Instagram’s advertising business is a fairly simple equation:

# of active users * % active that day * # of sessions per day * # of images viewed per session * # of ads displayed per image shown = # of ad impressions


# of ad impressions / 1000 * CPM = Revenue


# of ad impressions * click-through rate * CPC = Revenue

(Or some variation of these. Perhaps Instagram uses ads per sessions instead of ads displayed per image shown. Same with CPM; CPM could also be interpreted as eCPM based on click-through rate and CPC.)

So, if you’re looking to increase revenue, the impact from improving user retention and engagement are clear: active users, # of sessions, % daily active, all of those can be directly impacted by launching an algorithmic feed that delivers the images that people want to see. It’s also clear that a more engaged userbase could potentially click on more ads, increasing the performance of the ad units and delivering higher effective CPMs.

How much of an impact will it have? That’s a hard question to answer without seeing the data from their tests, or without knowing how it is being implemented. But with more than 400 MM users, small lifts in usage can still have a massive impact in absolute terms.

Decreased organic reach increases incentives to purchase advertising

The next set of impacts are more nuanced, but perhaps more important and more meaningful to Instagram revenue. Decreasing a brand’s organic reach on Instagram will decrease the amount of their posts that are shown organically, and increasing their need to purchase ads to generate the same reach on Instagram. If the lesson from Facebook holds on Instagram, attempts to build an organic follower base will lead to decreasing returns as the algorithm depresses organic reach.

But the impact doesn’t stop there: more brands needing to purchase reach will eventually impact CPCs. Since CPCs are a result of a marketplace of advertisers bidding for clicks to segments of users, more demand from advertisers will increase CPCs, as long as the increased demand outweighs the potential increased supply of clicks by users.

How much organic brand reach will brands on Instagram look to purchase from Instagram? Some might think that decreased brand reach could result in increased interest and dollars flowing to influencers, as brands could look to use influencers organic reach as a way to replace their own. But the dynamics there aren’t that simple. The metrics and measurement of ad performance simply won’t be as good for advertising dollars spent through influencers as it will be for dollars spent on ads, and the process for purchasing ads will be far simpler. Instagram and Facebook have sales teams calling on agencies and brands and building relationships with them; they have a product that makes it easy to spend money on Instagram ads if you’re already spending money on Facebook ads (really, it’s as easy as clicking a checkbox); they have metrics and measuring tools and APIs and an ability to report out details on how the ads perform that simply aren’t available to influencers. That’s not to say that this “kills” influencer marketing, I don’t think that’s the case. But it surely makes a brand (and an influencer, frankly) think more carefully about the right way to execute an influencer campaign.

And there’s also a big unknown here. Right now Instagram’s suite of ad products is fairly limited, and that’s largely determined by their product. An algorithmic feed can be executed in multiple ways, and as Instagram thinks of ways to implement algorithms they could also launch new ad products, with new pricing, new rules, and new results.

* * *

This is how I think about business: behind every business is an equation that takes a variety of inputs around how the product performs and what people do and pay, and out comes a set of revenues and costs that are tied to the operations of the business. Major strategic changes in the product are often tied to major business strategies, and while a financial model may not give an exact answer to a product question, it can inform and shape how to build products.

In Instagram’s case, the product rationale for the shift is based on a desire to build a better product for users, but there’s also a deep business strategy behind it that could have a fundamental impact on their advertising business. 2

  1. Full post at Algorithms and The Pursuit of Advertising, or a Snapchat-focused excerpt on Medium

  2. Want to model an advertising business? Use the new Foresight Advertising Model


In 5 free lessons I share what I’ve learned from helping 23,000+ early-stage entrepreneurs from 116+ countries learn financial modeling.

+ Get a free template cap table model detailing how investment rounds and exits work for founders and investors.