Incrementality measurement has been praised as the holy grail of advertising measurement. Some say that if you want to improve your measurement - you should understand incrementality.

Incrementality is the north star of marketing, as whatever is incremental is by definition - “adding value.”

 This article can help you understand the following:

• Measuring Impact: The Beginner’s Guide to Incrementality,
• What questions can I answer with incrementality measurement?
• Why do advertisers need incrementality for proper measurement?
• How to measure incrementality?

Regardless of your business or budget size, you need to understand the concept of incrementality.

What is incrementality ? 

Incrementality is the measure of the additional impact or lift generated by a marketing activity beyond what would have happened without it.

Incrementality measures the impact a marketing channel, campaign or change has on any KPI. One of the best things about an always-on incrementality measurement is that you do not have to stop any of your marketing activities to be able to measure their impact on one another or your KPIs. This is essential when measuring the incrementality of Facebook, where overlapping audiences and attribution challenges are common.

In the context of ads, it can be referred to as the measurement of ad effectiveness and it measures the true effectiveness of advertising activities irregardless of tracking.

The goal of paid advertising is to create incremental revenues. Whether if it is to establish a stronger brand equity, or to push people to complete a purchase, download an app or use it.

Incrementality in marketing is vital to distinguish between results that were truly driven by advertising and those that were not.

Incrementality testing requires Advertising to create various scenarios to isolate conversions data. Tests use changes in the marketing activities to compare how a change in activity influenced campaigns performance over time.

The goal for marketing in any organization is to drive growth by driving customers and prospects through the marketing funnel. Awareness > Interest > Desire > Action

Advertising efficiency is reached when the Advertising budget spend produces results that would not have happened if it was not for the Advertising activities.

Why do advertisers need incrementality ? 

Measuring incrementality is the best way to ensure that the sales results (attributed to paid marketing) are results that would not happen if it was not for the advertising activities. 

Without incrementality measurement, advertisers could be spending advertising budget continuously, believing that their advertising activities are producing incremental sales, while the reality could be that the activities are actually cannibalizing sales that would already happen without advertising. 

What is incrementality in marketing?

Incremental sales lift has been the goal of marketing for decades, and incrementality measurement has always been the goal of measurement. Incremental marketing is the key.

We set out to evolve digital marketing from the measurement of traffic to the measurement of value with INCRMNTAL.

Incrementality measurement until recently focused on experimentation. Segmenting audiences into a control group and showing those audiences with PSA or Ghost Ads, comparing the results of a campaign shown to the control group vs. the result of the general campaign. 

This approach usually produced biased or inconclusive results, as there was no ability to know if the control group was “clean” and unaffected by other campaigns running.

Various other attempts to test incrementality were done by blacking out advertising all together for a period of time - but this approach had such high opportunity costs and only provided conclusive results for the time the test was performed - that most advertisers abandoned the idea of performing such tests.

Our challenge at INCRMNTAL was: How would we know if a user was going to perform an action, even if they were not advertised to? 

The answer: we don’t

In essence, measured incrementality does not require experimentation. 

Learn More: Incrementality in Marketing

True attribution focuses on incrementality

Our initial idea was: we will build “better attribution”. We wanted to build an attribution solution based on 1st party data, and apply machine learning to understanding the multiple touch points a user has with ads. 

But this was a moot point - multi-touch is practically impossible in the mobile app ecosystem, as user data is becoming obsolete. 

We also figured that attempting to help developers by offering a new measurement SDK is not helping the developers. No one wants to integrate another SDK.

Our research, had us understand that developers are not in need of “better attribution” - attribution as it is - is ok. But attribution can lead to terrible outcomes.

Once we established a few ground rules, we had our direction

• We do not challenge attribution data
• We are not offering to replace attribution
• Incrementality testing is done in retrospect
• Incrementality measurement does not happen for a single user

How do we measure incrementality?

At INCRMNTAL, we’ve developed a groundbreaking methodology that eliminates the need for traditional experiments like A/B tests or randomized trials. Our approach is designed to measure marketing performance reliably, even in complex, noisy environments, using advanced automation and causal discovery techniques. Here’s how we do it:  

Step 1: Data Collection and Enrichment

• Retrieve all marketing activities: We automate the process of collecting marketing data from platforms. If platforms don’t support this directly, we infer activities from costs.  
• Enhance activities with external factors: To provide a fuller picture, we incorporate additional data, such as:  
  • Weather patterns  
  • Stock market trends  
  • Google search trends  
  • Holidays and seasonality factors  
• Score activities for importance: Each activity is assigned a dynamic score based on:  
  
  • Type and volume of the activity  
  • Timing and context  
  • Other entity-specific features  

Step 2: Predict Outcomes with Ensemble Time-Series Models 

• Build predictions using an ensemble of models: We combine the strengths of multiple time-series models (e.g., ARIMA, Prophet, and Causal Impact) to predict what would have happened without interventions.  
• Train and validate dynamically:
  • The model selects an optimal training window by analyzing data several weeks before the prediction date.  
  • Data is split into training and validation sets, with models scored for accuracy using metrics like MAPE (Mean Absolute Percentage Error).  
• Create a robust final prediction: By weighting the results of all models, we produce a final prediction wrapped in a 95% confidence interval, offering a reliable benchmark for comparison.  

Step 3: Understand Causation with a Distribution Algorithm

• Analyze overlapping activities: To account for multiple simultaneous marketing interventions, we use a causal discovery algorithm that distributes the observed effects across activities.  
• Incorporate contextual details: The algorithm factors in:  
  • Activity type and platform  
  • Adstock (carryover effects of advertising)  
  • Time and location of interventions  
• Leverage reinforcement learning: This dynamic system continually learns and updates expected behavior for each activity, ensuring accurate insights into short- and long-term effects.  

Step 4: Continuous Measurement and Insights 

• Measure continuously, not just at single points: Our continuous measurement algorithm evaluates marketing activities over time, providing a baseline to compare all contributions.  
• Calculate both marginal and absolute impact:  
  • Marginal values show incremental contributions.  
  • Absolute values reveal total impact.  
  • Refine predictions daily: The system adapts to new data and evolving marketing dynamics, offering consistent insights.  

The Outcome
With INCRMNTAL’s methodology there’s no need for disruptive A/B tests or experimental groups. Marketers could be measuring the incrementality of facebook, or get a clear view of what works (and what doesn’t) in real-time.  The system adapts dynamically to changes in data, ensuring accurate predictions and actionable insights.  

By combining automation, ensemble modeling, and causal discovery, we’re making marketing measurement smarter, faster, and more reliable.

Is incrementality measurement difficult?  

(short answer: no) 

The role of advertising within the marketing space is to attract customers, especially new customers, to engage with a brand, leading them within a 4 stage funnel. 

Advertising is most effective when it produces incremental results – i.e. sales that would not have happened if it was not for the advertising activities.

During the LUMA Digital Marketing Summit keynote presentation, LUMA presented the following slide saying: “True Attribution Focuses on Incrementality”

Marketing is a broad term describing a company’s activities to promote the sales of a product.

Each element in the 4 Marketing Ps may influence a successful marketing strategy:

• Product – the item being marketed
• Price – the price point (and model) for the product/service
• Place – access customers have to reach the product
• Promotion – Advertising activities to create awareness leading to action
  In this article, we wanted to focus on the promotion part (advertising), but that is not to say that the 3 other Ps do not matter!

In today’s competitive world, advertising activities are a necessary part of almost every product in the market. If you are a mobile app developer, you are competing against thousands of app companies from around the world. 

Media vendors have made advertising lucrative by offering performance pricing schemes where advertisers can sit back and pay conversions rather than risk the costs of media which may not lead to any conversions.

While it’s tempting to advertisers – performance pricing has one big caveat – it incentivizes media vendors to optimize their own media strategy towards the low-hanging fruit: Targeting users that may likely would have converted, thus reducing the media inventory that is required to be used.

The following graph shows that an advertiser may launch campaign activities with a new vendor. Based on reporting – the new vendor is providing great value and a positive ROI.

However, zooming out to see the bigger picture, shows that this new vendor does not contribute to the total volume of conversions, but is receiving credit away from the organic traffic.

This vendor may have a high overlap with organic users.

Understanding incrementality can be very simple, but may require a marketer to make a drastic move – stop all advertising activities for a while, to see what happens…. 😬 

Using the unique feature built by INCRMNTAL, marketers can use Time Machine to measure incrementality of past campaigns and actions. 

You may be surprised that large companies who stopped all advertising activities found that up to 80% of their advertising activities were redundant, sometimes including fraudulent publisher activities, and this was while using the latest mobile measurement and analytics technologies available.

Most marketers are unwilling to perform such a drastic step – as while saving is important – most marketers would rather unlock the value of their spend rather than “save it”. Or in other words, cut the right marketing spend. 

While the above example shows a clear picture of cannibalization, the reality for most marketing campaigns is more subtle.

An alternative solution to “Stop all Advertising” is available, by utilizing differences in different algorithms and examining the traffic mix on a narrow signal level. Utilizing algorithms allows a platform to continuously compare signals to signals on a combination level and point out opportunities for optimization based on incremental value while reducing cannibalization.  Here you can find a breakdown of the different techniques of incrementality. 

What incrementality isn't ?

Incrementality is not a replacement for attribution, nor a method to track clicks, impressions or conversions. Some ad networks offer incrementality tests, showing the marketer that their own incremental test proves that they produce incremental ROAS - but Return Over Ad Spend is not the correct measurement of Incrementality.

True incrementality is measured in total ROI - Return Over Investment. Making sure that the advertising activities are producing additional value. In other words, incrementality can help you identify where you are overspending and cut the right spend.

What questions can be answered with incrementality measurement ?

Incrementality measurement can help answer very important business questions, such as:

• What is the real value of my ad spend ?
• Is the advertising spend in media vendor X producing incremental sales ?
• Why did the performance of a certain campaign change without doing anything ?
• Did a price increase or a budget increase produce incremental sales ?
• Which campaigns are generating an incremental lift with other paid channels ?
• Are the advertising activities in media vendor X cannibalizing organic sales?
• How are your offline channels/campaigns performing? 
• What is the point of diminishing returns?

Incrementality measurement can be valuable across all platforms and verticals. 

Incrementality is one of the leading measurement metrics for growth marketers and we have seen several teams irrespective of industry find various use cases for incrementality measurement as it lets you measure exactly what you to measure in order to eliminate redundant spend. 

Some of the most common uses for the platform have been: 

• Measure cross-platform marketing 
• Measure offline channels 
• Measure external events, such as temperature, precipitation, transport strikes, national holidays and more! 
• Measure iOS campaigns, since INCRMNTAL does not require user level data
• Cut redundant spend 
• Unlock efficiencies 

Industries we have seen successfully use incrementality measurement: 

• Gaming
• Fintech 
• Travel/Mobility  
• Lifestyle 
• eCommerce & Retail

More over we have seen different teams use the platform for different uses. Most users fall within three main categories: 

1. Growth Marketers 
2. Marketing Data Analysts 
3. Data Scientists 

If you would like to learn more about the platform first hand, sign up for a free demo.