Amazon’s Q2 revenue will beat analyst estimates, but by how much?

Jumpshot data predicts that Prime membership growth will once again foil analyst estimates

On April 28, Amazon released its quarterly revenue and earnings for Q1 and blew past most analyst expectations, causing the stock to rise 12%. While many analysts’ consensus was $27.98 billion in revenue, Amazon topped out at $29.13 billion. Taking out revenue connected to their cloud service Amazon Web Services (AWS) — $2.57 billion — that number was $26.56 billion. Although most analysts and investors were clearly surprised by the results, the Jumpshot team celebrated for a different reason: we nailed it. Using our clickstream data from our 100-million user panel, we predicted Amazon would report $26.97 billion in revenue from non AWS products and services. That’s not just close — that’s almost more accurate than Amazon’s recommended products algorithm!

In Q2, Amazon will release its earnings on July 28. Using the same methodology, we decided to give it another shot. Based on the clickstream data from our user panel, we predict that Amazon will report total revenue (minus AWS) of $31.39 billion, representing a 47% increase from Q2 2015. Average analyst prediction is $29.54 billion (including AWS) which is a 38% increase YoY.

How did we do it?

Jumpshot analyzed its user panel and found the users who performed predictive events on Amazon, such as signups and cancellations to Amazon Prime and product purchases, in both North America and in select countries internationally. This analysis was then correlated to all of Amazon’s North American and International earnings releases since Q1 2014. Tracking actual purchases and sign-ups/cancellations to Prime in Jumpshot’s panel and applying a basic regression model, Jumpshot was able to closely predict future earnings as these are two very strong indicators of overall performance. Honestly, what better way to anticipate performance by a company selling goods than to look at the number of transactions themselves combined with subscription and attrition rates of its paid membership programs?

Looking at Jumpshot checkout data from the US and select international markets, we saw a very strong correlation to previous earnings releases.

We ran this same analysis for a number International markets, including Germany, France and Japan. Comparing past earnings reports with transaction and Prime data from the panel, Jumpshot’s data scientists created a regression model going back to Q1 2014. Using these coefficients and refining the formula, they were able to hit the nail not just on the head but smash it through the Wall Street bulletin board in one blow.

We are firm believers in the power of data-driven intelligence to predict earnings outputs with companies, especially eCommerce.

For more information about Jumpshot, how we were able to pull this off, or what else we can do, please contact us at

What is Jumpshot?

Jumpshot is a marketing analytics company that helps marketers understand their customer’s entire online lives. From the key sources of traffic to a site, to the browsing, consuming, and buying behavior on a site, to where customers go once they’ve left a site, our platform reveal the entire customer journey. Jumpshot tracks more than 160 billion monthly clicks from our 100-million customer panel’s clickstream activity. In short, we are able to see every single click that our user panel performs in the order that they do them from January 2014 through yesterday.

Like what you read? Give Deren Baker a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.