Since Facebook teamed up with Dunham, our goal has been to help advertisers understand the effectiveness of online and offline marketing campaigns for online and fast consumer products. Deng Hanbei analyzed 37 Placement Optimisation campaigns on Facebook, allowing us to study this issue for the first time through meta-analysis. This analytics provides cutting-edge insights and actionable data to help advertisers optimize their media strategies to advance the biggest sales impact of Facebook campaigns.

Through these 37 campaigns, Dunhambey found that their average brand sales increased by 3.7%¹, which means that the average household spends an average of ￡14.93 more²。 In the analysis, Deng Hanbei focuses on two sales factors:

1. Home penetration of goods or brands

2. Household consumption expenditure for goods or brands

Dunhambey found that household penetration and household consumption spending both increased after a marketing campaign came to light³, the average penetration rate increased by 3.8%, and household consumption expenditure increased by 0.9%.

To give marketers and planners more groundable insight, Dunhambe identified three main driving factors in sales marketing: reach, frequency, and cycle.

Last year Facebook published an article based on the U.S. CPG⁴Campaign reach articles that point out that campaigns with more than a quarter of coverage drive penetration and sales increments. Mr Dunhambey tested his data in the same way and found similarities in the UK market, although the details varied slightly.

Mr. Deng mimicked Facebook's analysis and combined with his own reality to segment existing campaigns by reaching more than 8 million people and less than 8 million people. While this is not the best indicator for broad positioning and precise positioning, the principle of how actual marketing campaigns are divided also reflects the power of positioning.

Focusing on marketing campaigns with below-average coverage (mostly focusing on a portion of the audience) produces a relatively high household penetration rate, while broad coverage campaigns generate relatively high household spending (2.6 times) and a high return on advertising (9.4 times)because these ads reach more people. This does not mean, however, that the same ad idea works for everyone, and Dunham's analysis shows that targeting as many people as possible in a similar group (using similar ad ideas) yields the greatest return on sales.

Another paper published last year showed that the effective frequency with which marketing campaigns are launched is determined by their business objectives. The article found thatMarketing campaigns for brand-building content are already showing significant results at low display frequencies (approximately 1.5 times per week) (increasing frequency does not result in additional improvements), but marketing campaigns with shopping appeal require more frequent impressions to achieve results (approximately 2 times per week)。 However, there is no "accurate" frequency level, and advertisers' marketing campaigns should take into account market and environmental factors that determine whether the effective ad display frequency should be high or low.

Mr. Deng reviews his data to understand the impact of ad serving frequency on campaign performance. Group campaigns based on frequency levels⁵Deng Hanbei found that higher display frequency marketing campaigns drive more household penetration, more household spending and drive greater sales growth in goods or brands.

In addition, when frequencies and cycles are combined to be more than or equal to 6 weeks and less than 6 weeks, we find that establishing a higher display frequency over a longer period of time can achieve the greatest sales increase.

Deng Hanbei's analysis shows thatAdvertisers plan their campaigns with a broad and relevant population that lasts more than six weeks and is shown once a week to achieve maximum sales returns in media programs, including Facebook。

If advertisers know that people other than focusing on people won't buy their products - perhaps because they're particularly directional - then precise targeting may be best, but the more general rule is that the maximum return on advertising can be achieved by targeting the widest possible relevant audience.

Comments:

[1] dunnhumby Sales Impact is built with an unexposed vs. exposed methodology.  Sales lift is the percentage uplift from the unexposed group to the exposed group.

[2] For comparability, wherever possible we look at uplift in percentage rather than value terms due to the differences seen between categories.  For instance, a 5% uplift in the food category will give a lower value uplift than a 5% uplift in household care, as the typical product costs less money.  For this reason, dunnhumby focuses on percentage uplift wherever possible, unless the value is needed to tell an important part of the story.

[3] dunnhumby measures campaigns run on Facebook, Instagram and via Placement Optimisation (across both Facebook and Instagram)

[4] Consumer Packaged Goods, similar to FMCG, or Fast Moving Consumer Goods, in the UK

[5] “Lower frequency”, or “f<1 pppw”, and “Higher frequency”, or “f≥1 pppw”

Download the report in English