What is and how to use it (+ examples)

I have worked in digital marketing for over ten years and I am always looking for ways to improve marketing analysis and reporting. Working on the leading digital agencies and my own multi-channel agency, in my work the most important is the modeling of Media Mix (MMM).

I also faced the SEO PPC agency with Leigh Buttrey, and we are dedicated to multi -channel marketing. She runs PPC, I run SEO, and we both know that our work affects another. We also work closely with the inner teams that have many other channels within their marketing media.

When it comes to marketing, I’m sure we can all agree that Omnichannel’s approach is the best.

Challenge? Well, that’s in the analysis. We all know that customers rarely pretend thanks to one medium, but what media is stimulating results?

Well, that’s where the Media Mix modeling is played.

In this article, I explore modeling Media Mix (mmm), starting with its definition, frame and examples of MMM action. I reached with merchants, sales experts and owners of companies using MMM in my marketing analyzes. They share real life anecdotes and tips that will help you feel confident about MMM.

Let’s go.

Content

If you understand how your channels “play” together and how influential each channel is, you can optimize your media planning. For example, you will better understand which channels we invest more in what purpose.

Following-I will share some real-life examples of how MMM transformed the business approach to marketing.

Modeling of the media mix in marketing

Source

Modeling Media Mix is ​​the best friend of marketing – although it is not a small feat to set up. You want a lot of clean data, in the ideal range of years.

MMM uses historical data to recognize and quantify relationships between marketing channels and their effect on business goals such as conversion or income. It is very different here from other models.

Models of the last touch attribution, for example, consider only the last marketing channel that resulted in sale, while Attribution from the first touch makes the opposite.

I feel that most merchants are happy to say that sales rarely come from one channel, and every media play a role.

In addition to modeling Media Mix, traders can predict future performance, which helps them make marketing decisions such as budget allocation.

With mmm, markers can:

• Collect historical data on things such as marketing consumption, sales, trends, etc.
• Develop a statistical model that explains the relationship between marketing activities and business outcomes.
• Interpret the results to understand the effectiveness of each marketing channel.
• Use insights to divert budget and resources for maximum ROI.
• Predict future performance based on various marketing scenarios.

In my opinion, the primary value of MMM lies in data. Instead of making decisions based on a sense of gut, you can quantify the role of canal in marketing. You will also move away from the measuring data on the downtown channel to help make wider decisions aimed at data with complete image of marketing performance.

Mix -model frame

Source

The Media Mix frame consists of six steps.

1. Data collection It relies on high quality, longitudinal data from different sources and marketing channels. As above in the picture, it could include sales, marketing consumption, consumers, production, economic and competitive information.
2. Data hygiene It is as simple as it sounds, but it is a long and very important step. Includes cleaning and withdrawal of data into a unique set of data ready for analysis. If you don’t understand that, you won’t get the correct data output. Spend your time here.
3. Model development It usually depends on machine learning models that will help you understand the relationship between marketing inputs and business outcomes.
4. Analysis It is best done with some human intervention. AI can do a lot of analysis and is amazing for analyzing large data sets, but marketing is very tin, and the human inspection of findings is necessary.
5. Optimization It greatly varies about acquired insights, but with your new insights based on data, you can optimize your marketing and budget distribution for future campaigns.
6. Forecasting is the way it sounds. Now that you have information, you can predict the potential outcomes of different marketing scenarios, create hypotheses, test them and repeat them until your marketing leads to the desired outcome.

Examples of modeling media mixes

The best way to hear about MMM and its impact is through real life examples. I was delighted with the incredible insights received from the merchant below.

SPOT synergy between channels

Aaron Whittaker is a vice president of a generation of demand in the Thrive internet marketing. Asked about MMM’s value, Whittaker said he “transformed how we award marketing budgets and measure the impact on the channel.”

One particularly valuable application that Whittaker found was to spot synergy between the channels. In this case solid use, Whittaker explains: “When analyzing the performance of the holiday campaign of retail client, instead of looking at the channels isolated, our mmm discovered unexpected synergy between radio advertising and social media.

We have found that radio ads during the morning trip have increased social media engagement for 25% in the coming hours – an insight that would not be visible through traditional models of attribution. “

What do I like in this: Marketing attribution is a huge challenge for any job, and without MMM, it is very easy to miss the value that Radio added in this case. It would be easy to assume that visits to social media, following, engagement, etc. What often happens is that efforts are fully attributed to social media, but the reality is that Radio has a role here.

With this information you can better rationalize why he did part of the media. Better, you can target radio media at the right time (in the morning, when it turns out to be effective).

For advice: If you work on your ads, fight to get the results you want or have proof (thanks to MMM!) To work with your advertisements for you, then see the HUBSPOT paid media. This makes the easy operation of organizing your media planning and buying media.

Quantify the long -term construction of the brand

Whittaker provided many examples for mmm. The choice that will join this article was a challenge! I had to include the value of a long -term brand construction and how MMM can help quantify its role.

On the construction of the brand, Whittaker says: “What is fascinating is how MMM helps quantify the long -term brand construction activities. We found that subcastation sponsorship showed minimal immediate ROI, but our modeling showed that it significantly contributed to reducing the cost of buying on other channels over six months . This insight has helped to justify the continuous investment in the brand awareness. “

What do I like in this: Similar to the upper point, I really like how mmm helps justify marketing efforts that might otherwise go unnoticed. It is true that if the channel does not result in immediate ROI, it becomes “frivolous” using the last touch attribution, but with MMM you can see how these media work for your business.

Understand a crossover between internet and off -hearted media activities

Peter O’Callaghan, head of marketing in Scragingbee, found the highest values ​​using MMM to detect regional trends. O’Callaghan describes mmm as transformation.

He says he helps “assign budgets, improve messages and identify growth options. It is a powerful tool to predict where to invest and where to withdraw. “

Asked for example, O’Callaghan says: “It helped us to determine California and Texas as a focal point for scraping demand for API, contributing to 40% of our guides. By diverting $ 5,000 in the campaigns targeted by Geo, we increased the regional engagement by 30%, shortening our sales cycle by almost two hours per lead. This regional focus continues to shape the way we approach the campaign strategy. “

Tips for using Modeling Media Mix

Tip 1: Start your analysis with a lot of data.

With its winning examples of MMM, Aaron Whittaker advised that anyone who starts with MMM analysis should start with “at least 18 to 24 months of data”.

The more information is, the easier it is to see trends. Whittaker explains, “18-24 months of data (help) calculate seasonal patterns and long-term effects. We have found that shorter periods of time often lead to the wrong conclusions on the efficiency of the channel. “

Whittaker has another mmm example that shows the value of the data beautifully.

“A surprising discovery appeared during modeling seasonal influences. Our analysis has shown that the effectiveness of different channels has dramatically varied until the season. Marketing through the E -šte reached its peak during the winter months, while the outdoor advertising was delivered to the highest ROI during the summer. This has led us to develop dynamic budget distribution strategies that change consumption based on seasonal efficiency. “

What do I like in this: I am sure that many merchants who read this head in agreement. We all know that we need the data – and the more, the better – to make an appropriate analysis.

Tip 2: Make sure your data is clean.

Peter O’Callaghan advises that “MMM works best when you have clear, measurable goals. Without defined outcomes, it is easy to misinterpret insights and affect incomplete information.”

It is easy to miss where your data should work, but O’Callaghan has several tips for that.

• Watch out for poor segmentation. O’Callaghan explains that bad segmentation hides valuable patterns. He says: “If the data is too generalized, key trends that distinguish between users’ groups can be lost. Punching data in smaller, meaningful segments allows you to understand unique behaviors and preferences of different audiences.”
• Balance the assessment of short -term trends or seasonal spikes. O’Callaghan warns about short -term trends and seasonal spikes, explaining: “Mmm Exits can occasionally delight if you are too much short -term trends. Seasonal spikes or external factors can distort the results if you do not count them. For example, a one -time traffic lean has led us to overbid the performance Email uses mmm results with long-term measuring data to ensure a balanced display. “

What do I like in this: This is the feeling we heard earlier in this article. I like that O’Callaghan recommends cross -results with long -term measuring data. This advice is perfectly aligned with Aaron Whittaker’s advice on the launch of MMM with long-term data.

Start with Media Mix Modeling

As a seller and primarily SEO, I know the value of modeling Media Mix. Still, writing this article and talking to other merchants, I see MMM help companies to make better marketing decisions. Instead of feeling Like certain marketing media for you, MMM helps you prove it.

So, if you want to start modeling Media Mix, do it. Remember to collect these long -term data and cross shorts with long -term trends.