Fred Askham headshotMetrics make the marketing world go around — and these days, there’s more data than ever to make it spin. It doesn’t take an advanced analytics company to transform data into meaningful metrics that can make an organizational impact, but it definitely helps. That’s why we turned to our frequent data analytics consulting partner, Fred Ashkam, for his best advice on making good use of business intelligence.

Fred is the director of analytics at IMM, a solutions-focused marketing agency that believes in the power of data to drive business decision making, which makes him the perfect person to field our five most pressing questions about the business of business analytics and reporting.

Everyone loves data, but many struggle to translate insights into action. As someone with a strong business analytics services background, what advice do you have for them?

Here are three best practices to consider for building the framework that will guide your business analytics and reporting:

1. Create data accountability.
In process areas where there’s a large amount of SME decision making taking place, such as media buying, it’s important to understand what’s happening and why. Answering those questions takes more than just data — it takes quality data. Insights are only as good as the data they come from, and poor quality data can stall your business from taking effective action against its data. For example, you can’t expect to make solid decisions if your naming standards are poor. Or if your data is disaggregated across multiple systems. Or if you simply aren’t measuring an important decision-making factor.

Building a solid data framework and creating data accountability across your organization is one of the first steps to being able to use data to make actionable decisions. Consider setting up an independent analytics database to help take the burden of analytics processing off your production level systems and to allow your teams to more effectively aggregate, query, and report against the data you have available.

2. Avoid the trap of thinking more complexity equals better results.
Every data scientist thinks that a recursive ten-layer neutral net is super cool, but that doesn’t mean it will produce better results than a simple logistic regression. It is incredibly important for organizations to make decisions based on data with the strongest predictive power — and that isn’t always the data with the fanciest sounding name.

Think about it this way: Would you rather have an AI-driven ML system for bidding on media in 2nd price auctions or use the simple, well-defined bid-to-value strategy? The second, right? Of course. Results should always win over flashiness. (Plus, this strategy also proved to be the best-performing method of bidding in second price environments by John Forbes Nash, the Nobel Prize-winning mathematician who inspired A Beautiful Mind.)

To that end, you should always assess your systems based on their predictive power, not their in-set correlative values. When it comes to model assessment, always remember that predictive power is more important than R Squared — and don’t confuse the two, or you will certainly pay the price of using an overfit model.

3. Be patient.
Data is fuel for a value-generating process — not a silver bullet. Most people expect data to produce some large insight that creates a huge amount of value quickly. The reality is that the value of data-driven decision making is that it allows you to make consistently better choices over time. Some of the value-add from some of those optimizations is small, which may make it seem mundane, but the sum total of its impact can be significant over time.

What metrics are too often overlooked or undervalued?

Incremental return on ad spend (iROAS) is probably the most undervalued metric in the media space. It provides a metric on the return from media that is directly contributed by marketing, and it helps marketers avoid the pitfall of advertising to consumers who would have already purchased without the influence of marketing.

In terms of business intelligence, Consumer Lifetime Value (CLTV) is often the most undervalued metric that brands have. Not only does it provide a key metric for segmenting out consumers to help better identify key areas of growth opportunity, it’s also critical for understanding the level of investment that should be put against acquiring new leads.

Businesses often fail to generate appropriate growth because they are underinvested in new business generation or consumer acquisition. Understanding the long-term value of a consumer or new business relationship can help them justify the appropriate investment levels to their boards to ensure sustainable long-term growth.

What trends do you anticipate in 2021?

Try to predict anything in 2020… A meteor maybe? Extra-large locusts that eat meat for some reason? All kidding aside, here are three trends I expect to continue to develop in business analytics and reporting in 2021:

Cloud dominance: The cloud is here and it’s not going anywhere. It provides cheap, scalable storage and processing that’s super helpful to businesses, hence its widespread adoption. If you don’t have the skills or expertise to bring cloud solutions to your business, partnering with an analytics consulting company could be a smart investment. But be sure to give yourself ample time to migrate to a new system — the cloud is great, but cloud migration isn’t always a simple task.

Data privacy: A tidal wave of regulatory and technological changes, including recent updates to Apple’s iOS14 privacy practices as well as Google’s plans to eliminate cookies on its Chrome browser in 2022, have made privacy one of the top issues facing digital marketers today. These developments are going to have a significant impact on eCommerce and marketing data.

You will likely need site updates to comply with new cookie policies to maintain cookie-based functionality. There are also specific server-to-server integrations with media platforms like Facebook, which must be set up to maintain marketing performance. And you’re going to need to enact a host of procedures to deal with data deletion for consumers who demand it.

Expansion of data accessibility within organizations: Reporting tools and platforms have made great strides forward when it comes to turning databases into user-friendly dashboards. They can also effectively take the technical knowledge out of the task of querying a database. This trend will likely amplify in 2021 as more companies move towards self-service business intelligence (BI) models.

In this type of BI model, a data analyst builds a highly flexible reporting dashboard and then provides access to it to the non-technical SMEs. The people closest to the data can then use the dashboard to answer their own questions instead of filtering them through analysts who may not otherwise understand the business context. Self-serve BI dashboards are becoming an increasingly popular request from clients who are interested in business analytics services.

What is the most surprising insight you’ve gleaned from a business analytics and reporting effort?

I’ve found that people who eat at casual dining establishments are much more likely than the average individual to purchase regenerative hair growth treatments. That was a surprise!

On a more serious note, at IMM, we once uncovered some interesting insights around a brand’s consumer retention rates. This brand should have had a strong following of champions — it had been built around brand loyalists, after all — but it had a customer retention problem. They were surprised (and alarmed) to discover the issue, but it helped them craft new strategies for retention.

We also once proved that a brand was wasting a good amount of its media spend on retargeting media buys. They thought it was a profitable marketing buy, but our suggestion that they pull back on spending in that area significantly increased performance and kept sales volumes level with decreased spends. That meant they were able to increase their spend in other areas and boost new customer acquisition on other channels. In this case, data analytics consulting led to the brand’s best eCommerce year to date, even before the pandemic hit.

What do most people misunderstand about paid media?

There are so many misunderstandings about the proper metrics to evaluate paid media effectiveness. Most brands use Return on Ad Spend (ROAS), but it’s actually a pretty terrible metric because it fails to account for incrementality. There’s a reason business analysts call ROAS metrics the comfortable lie. It’s comfortable because it’s the metric that’s always been reported, but it’s a lie because it’s demonstrably false — they show really high marketing returns on paid media, but brands don’t really see those returns.

My best advice for brand managers and marketing directors? Look into the myriad of fast, reasonably priced incrementality testing solutions on the market. You can learn a lot about what is and isn’t working in your media practice and quickly align the results with business goals.

We can help you find opportunities within your marketing and business analytics — and activate them. Learn more about what we do.