How to evaluate which half of your advertising spend is the most effective

Deepak Chandran, Zarak Khan / September 2020

Deepak Chandran, Head of YouTube Ads Research at Google APAC, and Zarak Khan, Regional Product Lead - Brand Measurement at Google APAC, share insights on the importance of measuring sales impact and explain how adopting a mix of measurement strategies can help advertisers monitor campaign effectiveness and make the most of their media spend.

Most marketers have heard the old saying: “Half the money I spend on advertising is wasted; the trouble is, I don't know which half.” For many, the answer to this question is as mysterious as the origins of the quote itself.

But here’s what we all know for sure: Brand marketers need to evaluate whether their campaigns are effective in generating incremental sales. Then they must understand whether these campaigns are efficient. In other words, do they generate enough incremental sales to cover the costs of the campaign so they can determine whether to continue or change course.

However, accurately measuring the impact of a marketing campaign — then identifying ways to improve the return on that investment — has only become more challenging as brands compete for customer attention in an increasingly crowded landscape of fragmented media channels and brand messages. What’s a marketer to do?

Brands that adopt a measurement strategy combining credible, comprehensive, and actionable solutions will, more often than not, find further opportunities for growth.

Well, having an informed strategy is a good place to start. There are two commonly used approaches to measuring the sales impact of media platforms. Sales lift studies, which measure the impact of a single platform, and marketing mix modeling (MMM), which looks at the impact of multiple platforms at once.

To help marketers better navigate their measurement challenges, we recently partnered with Nielsen to identify YouTube’s sales impact across APAC countries and uncovered some useful ways marketers can apply sales lift studies and MMM to determine the impact of their own brand campaigns.

Find the “working half” of your ad spend

Over the last three years, we ran 57 sales lift studies in APAC — spread across 16 categories and 46 different brands — to understand YouTube’s effectiveness.

Using a Nielsen Matched Panel Analysis study, we measured the impact of a single YouTube campaign while controlling for possible variables such as price, distribution, promotions, and other activities. This causal impact design allowed us to isolate the true impact of YouTube by comparing the sales between markets where we ran the campaign and control markets where we did not run any YouTube ads.

“The studies allowed us to measure the impact of YouTube in a variety of conditions while also allowing us to assess the effects that different YouTube formats and target audiences had on sales,” said Jayati Basu, Director - Marketing Effectiveness at Nielsen.

An analysis of these studies demonstrated three key aspects of our platform1:

YT sales impact_Inline Nugget_Inline Nugget #1 (1).png

“Partnering with Mindshare & Google enabled us to execute the Lays Smile Initiative exceptionally well. An end-to-end project encapsulating insights, creatives, media, and measurement culminated in a sales lift study where we saw YouTube deliver a 12.5%2 sales lift with a quality return on advertising spend (RoAS) as well,” shared Shailaja Joshi, Associate Director - Marketing, PepsiCo India.

Tips for advertisers: Embrace the power of experimentation. Adopt a sales lift study that drills deeper into which type of creative, audience targeting, or ad formats lead to the most successful outcomes. Additionally, continue to experiment to find approaches that work — and, equally, ones that don’t work — to ensure future campaign decisions can be based on proven results.

Monitor what’s working across different channels and get granular

By understanding which channels in your media mix are performing the strongest, you can optimize your media plans appropriately for maximum sales impact. MMMs allow you to assess the sales contribution of all the channels where your brand has invested over a period of time.

As an advertiser, breaking down digital media into manageable units allows you to form a clearer understanding of what’s working well and what could be tweaked to perform better.

YT sales impact_Inline Nugget_Inline Nugget #2 (1) (2).png

In our YouTube effectiveness journey, we wanted to examine how YouTube performs compared to other channels and platforms. We conducted another meta-analysis of MMMs that Nielsen had built over the last few years across APAC and compared YouTube to TV and other digital channels.

In India, we analyzed more than 100 MMMs from 2016 to 2018 and noted the following3:

YT sales impact_Inline Nugget_Inline Nugget #3 (1).png

We conducted a similar analysis in Japan from 2017 to 2018 and learned the following4:

YT sales impact_Inline Nugget_Inline Nugget #4.png

Tips for advertisers: Run MMMs to understand how your overall channel mix is working in regard to driving incremental sales for your brand. Focus on both effectiveness (in other words, incremental sales per impression) and efficiency (in other words, incremental sales per dollar) while making your media allocations. Also, consider working with a Google Measurement Partner to discover granular campaign-level YouTube/Search/Display data that you can use to understand the return on your investment so you can improve what’s working and fix what’s not.


Deepak Chandran

Head of YouTube Ads Research, Google APAC

Zarak Khan

Regional Product Lead - Brand Measurement, Google APAC

Sources (3)

1,2 Commissioned Nielsen Matched Panel Analysis for Google. Base = 57 YouTube campaigns measured during 2017–2020 (8 in Australia, 7 in India, 16 in Indonesia, 9 in Malaysia, 17 in Thailand). Count based on tested strategies with significant lift based on a one-sided 80% confidence interval. The average lift percentage and average return on ad spend is calculated as a straight average across studies. The ANCOVA methodology was used in all markets, except for seven studies in India, which used the time-based regression methodology to overcome data limitations.

3 Commissioned Nielsen MMM meta-analysis, 2016–2018. Base: A list of studies selected and compiled by Nielsen, consisting of all available CPG studies completed from 2016 to 2018 that contain YouTube, Other Digital, and TV results. This list includes 108 studies for India Brands across the Total CPG category. Incremental offline retail sales measured for TV represents the average across all TV broadcasters. ROI is Retail ROI, defined as total incremental sales divided by total media spend. The ROI numbers in the claim are spend-weighted averages across all studies for each channel.

4 Commissioned Nielsen MMM meta-analysis, 2016–2018. Base: A list of studies selected and compiled by Nielsen, consisting of all available CPG studies completed from 2017 to 2018 that contain YouTube, Other Digital, and TV results. This list includes 33 studies for Japan Brands across all CPG categories. Effectiveness is defined as total incremental sales per 1,000 impressions (in USD). Studies were excluded if either channel had fewer than 500 impressions.

The big picture: Optimizing ad spend to reach audiences that matter