Its Hot Grows Profit With Omnitail

Google Shopping

BACKGROUND, a leading retailer of exquisite diamonds, fine jewelry and diamond watches, struggled to drive profit through Google Shopping and paid search ads. A large number of fraudulent purchases skewed conversion data, making management of the program a unique challenge.

They came to Omnitail to develop a more accurate, profitable process for tracking revenue and managing their advertising. 

Increase in Valid Revenue
Increase in Valid Orders
Increase in Average Order Value
Previous slide
Next slide


Increase in Valid Revenue 


Increase in Valid Orders 


Increase in Average Order Value 


We investigated the account and discovered four main issues to address in our Google Shopping strategy: fraud, offline orders, inefficiency in the paid search program, and a long sales cycle. 

1. Fraudulent Purchases 

Fraud accounted for a large portion of daily purchases from This made reporting a nightmare: how do we account for the false revenue generated by fraudulent transactions? We needed to remove these values in as close to real-time as possible, to ensure the bogus dollars don’t affect bidding and other strategy decisions. 

2. Manual Orders 

Like other retailers of high-ticket or luxury items, also recorded a large number of transactions manually (from orders made over the phone, for example.) The problem was the inverse of the fraud issue. We needed to include these orders in a timely manner, so they could be included as factors for bidding. 

3. Paid Search Profitability 

The client’s paid search program was running in the red every month. The goal was to increase spend to bring in additional revenue and profit, but we needed first to fix the issues present in existing campaigns. 

4. Long Sales Cycle 

 The high-dollar value of’s product line also meant a much longer sales cycle than many other ecommerce retailers. Because Google usually operates on a shorter attribution cycle (30 days), this can make it difficult to attribute sales to campaigns. We needed to use an attribution model that would better represent the business. 

Omnitail's Advertising Strategy

Tracking and Removing Fraud for

We worked with the client to automate fraud reporting for use in our reporting and bidding systems. Any fraudulent orders were removed from the bidding process, so products with a high fraud rate wouldn’t receive huge bumps in spend improperly. 

We also include fraudulent order and revenue metrics in our KPI reporting, to track the rate of fraud over time. 

A New Attribution Model 

There was often a long research or consideration phase for’s products. As a result, we pivoted from a 30-day attribution model, in which sales were only attributed to an ad if a customer clicked on an ad and completed a transaction within 30 days, to a 60-day model. This allowed twice as much time for the customer to purchase as the previous model, and offered much more actionable data than the previous model. 

Making Paid Search Profitable 

We identified a number of keywords that were drawing a lot of spend with little or no revenue to show for it, and either cut that spend or redirected it to keywords that were driving sales. Within the first few months, the paid search program not only cut spend significantly, but achieved profitability with massive increases in valid revenue attributable to the program. We were then able to increase spend, which made higher-ticket items much more competitive in the marketplace. 

Screen Shot 2021-05-04 at 11.28.50 AM

Including Manual Orders 

The client also provided us with a daily count of manual orders and revenue attributable to those orders. We automatically synced this information with our bidding and reporting systems so those orders could be accounted for when adjusting bids or running ad hoc analyses (like gap analyses.) We could also then track manual orders over time, to establish trends. 


In the first year after signing on with Omnitail, saw a major improvement in profit and revenue. Identifying and removing fraudulent orders from their PPC analytics before they could impact bidding resulted in a 255% increase in valid orders, and a 230% increase in revenue.

Furthermore, average order value during this time grew 16%. The SEM program achieved profitability, and we were able to increase investment in the program to continue to grow sales. 


Increase in Valid Orders 

Ready to See Results Like This?

More Happy Clients