Tomi.ai
Scaling Profitable Customer Acquisition with Predictive Marketing

Consistently acquiring new customers without losing money is a challenge for any business. But it is a much more arduous task for LeadGen-focused and product-led businesses like B2B, SaaS, financial services, real estate, automotive, and many others.

The problem is the long sales or lead-to revenue (or trial-to-LTV) cycle. Because of it, most "long cycle" businesses have no choice but to optimize their digital acquisition campaigns for mid-funnel outcomes like leads or free trials. Not bottom-of-funnel business results like net new revenue or customer lifetime value every business is after.

This was precisely the problem Konstantin Bayandin - Founder & CEO of Tomi.ai - faced when he joined Compass as a Senior Director of Digital Marketing & Marketing Technology back in 2018. By then, he was a long-time e-commerce CMO who pioneered Predictive Targeting back in 2015 and paved the way for OZON – his previous company – to IPO at NASDAQ at a $10+ Billion market cap. Essentially, he all but forgot what it is like to buy leads and clicks instead of revenue.

A theoretical physicist by training and a marketing scientist by profession, he started thinking about a possible data science based solution. He discovered that website behavioral patterns if tracked at a sufficiently granular level (much more granular than is available in systems like Google Analytics or Adobe Analytics), can be used to reveal the true purchase intent.

At the height of the pandemic – in April 2020 – Tomi.ai was born with a mission to connect (primarily long cycle) businesses that solve life's biggest problems (like buying a house, acquiring a new profession, or growing a business) with their customers in a smoother, less wasteful, and more meaningful way.

Tomi.ai works by connecting "ultra-HD" behavioral signals to future CRM outcomes like revenue, LTV, or qualified opportunities with the power of machine learning. Their Patent pending pixel collects hit-level behavioral data. Moreover, their clickstream is almost as data-rich as a session video. The pixel is small in code and runs asynchronously not to impact the page load speed.

Tomi.ai then matches first-party behavioral data with the sales data from the CRM to train machine-learning models that predict the probability of a purchase and likely revenue or LTV. Finally, the predicted revenue calculated in real-time for each session is used for better targeting, bidding optimization for LTV or other value-based metrics, forward looking attribution, and predictive lead scoring.

"With our solutions, businesses with long sales cycles can increase their digital marketing ROI twofold with the help of predictive targeting," states Bayandin, Founder and CEO of Tomi.ai.

To date, the company has helped multiple banks and B2B SaaS customers to cut down CAC by 25-45 percent without losing sales and grow revenue per advertising dollar by up to 5x. For instance, one of Tomi.ai's customers, a consumer bank, faced the problem of high customer acquisition cost and relatively low quality of customers acquired from digital channels for their two flagship products: cash loans and credit cards. In addition, the digital marketing team struggled with making non-brand search a working and scalable channel that could generate new demand at a reasonable price.

Tomi.ai's solution to this challenge was leveraging predictive targeting and bidding optimization to optimize advertising campaigns for issued loans and cards, not mere applications. To make it work, Tomi.ai trained their machine learning model (on first-party behavioral website data and sales data from the customer's CRM) to score every website visit for the probability of a purchase.

Following a two-month A/B test, the customer experienced a 54% reduction in cost per issued card and a 24% reduction in cost per issued loan, as well as more than a twofold (x2.1) conversion rate increase for credit cards issued and an x1.5 conversion rate increase for cash loans. As a result, the customer could make the non-brand search a working channel for new-to-bank customer acquisition.

In a nutshell, be it businesses with a narrow target audience, large and variable transactions, significant offline components, or long sales cycles, Tomi.ai's platform provides the highest efficiency gains to all. By turning customers' websites into e-commerce from an Ad Platform perspective, the company allows customers to unlock smart bidding, help the Ad platforms train 10-30x faster, and enable 10x faster and cheaper A/B tests.