A serial entrepreneur who had founded four companies and had two successful exits, Improvado CEO, Daniel Kravtsov, learned after hundreds of meetings with digital media agencies and businesses, that all of them found it painful to monitor and report on digital ad campaigns.
Daniel’s reporting products and style, aggregates data across all marketing campaigns into one centralized dashboard. It has been such a saught after service, it pushed Daniel to created what is now Improvado.io – a tool that helps enterprises and ad agencies aggregate their marketing data in one place, in real time.
M.R. Rangaswami: How should companies think and calculate Return on Ad Spend (ROAS) and Customer Acquisition Cost (CAC), that would give them a deeper understanding of their complexity help companies increase their revenue?
Daniel Kravtsov: Each company needs to decide how it will calculate ROAS. One company may base ROAS on all Net New Revenue. Another might base it on Net New Marketing Revenue. A third might focus on the marginal fraction of Net New Marketing Revenue minus prime costs (which a company will spend anyway while creating a product).
Then the company has to get clear on how it calculates CAC. If a prospect sees your ad, for example, then reads an article mentioning your company, then reads a blog post, and finally converts through a paid organic click, each of those touches has a cost. One of the touches represented a direct conversion, but if you turned off all the campaigns except the one that directly converted, you would likely experience a big drop in conversion.
The modern data stack was created to help execute and understand such omnichannel campaigns.
In general, marketers ought to spend their digital ad dollars more shrewdly in order to create a more consistent return on ad spend. Here are some of the reasons they don’t do this:
- Growing data complexity. Some departments have more tools than employees. In order to gain a true understanding of Return on Ad Spend, or ROAS, the company has to fetch, normalize, format and combine the data from each of those tools. Few companies invest the resources to do that.
- Unattributed data and analytics misconfigurations. If your web analytics are broken or degraded, you won’t be able to properly connect leads directly with ad metrics.
- The rise of the modern data stack. It can be convenient to use CRMs to attribute marketing data. However, to do this effectively requires the routing of countless data sources into a CRM, followed by a lot of attention from data engineers and analysts. Moreover, the inflow of marketing metrics offers conflicting indicators for a single entry, thus demanding an external system to pre-process and harmonize it. We’re in the era of omnichannel, where a decent lead comes only after a dozen or so marketing touches. That fact further increases the complexity of such a setup, making it more challenging to assess the Customer Acquisition Cost.
Companies need to acknowledge these complexities in order to address them.
M.R.: Marketers often look at the number of touchpoints it takes to convert a customer’s early interest to revenue. How can marketers develop a greater understanding of that customer journey, and focus their marketing efforts to be less expensive?
D.K.: Fortunately, for the first time, revenue attribution models show marketers exactly how much it costs to bring prospective customers to each point of the journey, from awareness to purchase. Such models, when built properly, can show where the revenue comes from.
In a three-step customer journey, for example, we can measure first-touch attribution, last-touch attribution, and time-decay/reverse time-decay, in which we attribute revenue to steps that vary depending on when they occur.
Omnichannel attribution enables marketers to look at the data from different angles and to experiment, taking those angles into account.
However, implementation is not at all simple. You have to record events across different channels, collect them in a single storage location (data warehouses are the new nerve centers), normalize the data across channels, and then attribute revenue to channels based on the occurrences of similar touches in the customer journey of your prospects in order to determine which ones are likely to become ideal customers.
Global marketers, in particular, should study these models as they examine the impact of their cross-regional marketing, looking for similarities and differences in the behaviors of their ideal-customer cohorts. There’s no substitute for understanding how each ad contributes to the bottom line.
As part of this process, data warehouses have become the new nerve center of marketing organizations. Rather than scattered information across multiple tools, they serve as a central data storage, making it easier to connect vast amounts of data to BI and Analytics tools for further visualization and analysis.
Yet, data warehouses alone aren’t a silver bullet to cut on the data complexity. It’s not about piling up data at a central (or cloud) location — it’s about how you make use of that data.
M.R.: What would be the optimal pace of launching new experiments, and how can marketers ensure they have the right mix of short-term and long-term experiments?
D.K.: In our work with companies, agencies and brands, the biggest predictor of campaign success is the use of rapid experimentation and real-time measurement.
What determines whether you’re going to get the most out of a channel or not is directly connected to the frequency of your experiments, and how well you can measure them to synthesize the insights.
Iteration matters more than your budget, your time or even your creative assets.
If you want to get the most out of your investment in each ad channel of your choice and control for variables (for example, which message, which targeting, which time of day, which creative and which tactics worked best), you can’t just stop there; you have to dig deeper.
You need to be able to ask more detailed questions, compare different segments and deploy multiples of the same campaign, with slightly varied parameters.
The most successful companies and marketers launch new campaigns daily. Or more precisely, they launch new experiments every day. And the more experiments they launch, the more they learn about what works and why.
Although it may sound funny, the best way to ensure that you have the right mix of short-term and long-term experiments is to … experiment with your mix!
M.R. Rangaswami is the Co-Founder of Sandhill.com