An established eCommerce company in the renewable energy space was generating about $8M a year on their website.  They were using several channels domestically and internationally, including Amazon and retail.  The CEO wanted their e-Commerce website sales to grow significantly, but the team wasn't sure where to start.

The model

We began by building a channel model for the company based on their performance over the previous 12 months.  This model included traffic, conversion rates and sales by channel based on last click attribution so we could understand where sales were coming from.  We also created a baseline forecast based on historic patterns.  Once we had done this, we were able to identify the 'hole' between this baseline forecast and the goals of the CEO.

Its all about channels

Our model showed that organic and paid search were the primary channels, with most of the company's revenue coming from organic traffic.  We started with a Cost Per Acquisition (CPA)/Lifetime Value (LTV) analysis to figure out how much a customer will spend over their lifetime and what it costs to generate a new customer.  

We concluded that the existing Pay Per Click (PPC) agency wasn't doing much to drive sales or optimize marketing spend and that the Search Engine Optimization (SEO) strategy needed help with content.  Our next step was to introduce the client to new SEO and PPC firms. This new PPC company was able to increase spend significantly while maintaining the CPA targets.  

A new content strategy led to growth in organic search in new areas and markets over time.

Paid Social

We started a paid social test with a $10K per week spend.   This test generated a Return on Ad Spend (ROAS) of more than 5.  The spend was increased until the ROAS was around 4.  Paid social advertising because a significant contributor to growth over time as we were able to scale up spend.

Affiliates

Since this company was in the renewable energy space, it lent itself well to affiliates as education was needed to use their product.  We introduced the client to an affiliate management firm, and within 6 months the client was generating more than $200k per month from affiliates.  This was one of the most rapid affiliate ramp ups I have ever seen and this channel quickly became a significant part of the company's growth.

The Updated Model

Once we added these new channels to the model we monitored their performance and were able to project revenue accurately for future quarters.  However, not all of our tests worked.  

What didn't work

Throughout the process, we tested influencers, podcasts, and even ran radio tests.  None of these channels worked or showed potential. While this was disappointing, it is common.  Not every channel will work with every product and you can't know what does work until you test it.  

For example, in retrospect, we believe that radio did not work because the target market was a small percentage of the overall population.  This led to very high costs per new customer even with targeted ads and networks.  Influencers and podcasts appeared to be a potential solution to this because of their highly targeted audiences, but their reach was too small to have any significant impact on the company's sales and the costs were very high on a CPM (cost per thousand impression) basis.

The end state

After running through these tests and channel additions, annual revenue went from about $8M per year to more than $32M per year.  Every step along the way we updated the baseline model, identified new opportunities for customer acquisition, tested and progressed from there.  Not everything, or even most things, worked, but the ones that did made a meaningful difference.  A great reminder that you don't need to bat 1000 - or even 300 - to be rockstar in marketing.

Mike Morris is the founder of Kettle Hole Partners. When he is not trying to figure out mathematical models for marketing and customer acquisition he is probably riding his bike.