How can you unlock the greatest business value for your data?
Source s.D. Google
Looking back on 2019, "going out to sea" has become a hot word for the industry."Cross-border marketing" has become a new demand for many enterprises.
As one of the important platforms of overseas marketing, YouTube has certain research value to enterprises. The article shared today is one of two problems that marketers have in YourTube's campaign, and if you don't realize it in the marketing, you may miss out on your potential users.
75,000 Global YouTube Campaigns
For a long time, marketers have heard of it: relying solely on basic data like demographics(age, gender, etc.)The group portrait obtained is one-dimensional plane, not enough three-dimensional comprehensive.
But in the actual process, will still roughly simple choice of age, gender and other information to advertise. The end result could be the release of an ad with a potential risk of not attracting users, and missing out on potential users.
Granular user data can bring greater benefits.
For this, we can sense it intuitively.
Imagine at work, you know someone planning a family trip, buying a new phone, loving extreme sports, wanting to eat more healthy food, and so on. The more information you have, the more you know about him and the more certain you know how he feels about the product or brand. One day, there was a chance that you would have a chat the next day. At this time, you will have a feeling or expectation: I will run ads, do content marketing, the effect will certainly improve.
But relative to intuition, data tends to be more powerful.
75,000 Global YouTube Campaigns
How do I verify it?
For10 industries(including automotive, retail and travel)UseGoogle Brand Promotion（Google Brand Lift）Two-year summary and anonymous performance data for 75,000 global YouTube campaigns studiedThese data can measure the impact of video advertising on metrics such as brand awareness, ad recall, and consideration.
Start by dividing these campaigns into two categories:One category is to use only basic demographics to attract potential customers(e.g. ads specifically for women aged 25-34)；The other class uses what you call "advanced audiences" on YouTube.
The difference between the two is:The first type of user data used is very rough;The second type of user data is more granular.A way to attract users based on interest(E.g., the videos they watch on YouTube may speculate on whether they are foodies, travel enthusiasts, beauty experts)。For these pre-established groups,(have your own interests), call it a affable audience.
A way to create an audience based on the characteristics of your brand(For example, non-dairy brands may want to attract not only foodies, but also vegetarians)。This category is called custom interest-like audiences.
There is also a high-level audience for users who are working on certain products or services, which is called a marketable audience.
Second, determine the benchmark elevation of the indicators in each vertical industry.For example, on average, how much of the technology used in the sample increases brand awareness?
Then, to understand the effectiveness of not just using basic attribute data to reach people, marketing campaigns that use advanced audience features are further isolated to see how they improve on marketing metrics compared to the industry average.
What will you see?
The result is:
In all vertical industries, use campaigns from premium audiences for a variety of marketing goalsthere has been an improvement.
Use market audiences（From the perspective of user needs）To touch theFinancialCampaigns, users consider purchasing power increased 1.5 times, retail campaigns, purchase will increase d'or G.
Use Affinity Audiences（From user interest）Reached Telecommunications CampaignsAd recall rate increased 1.3 times。
Audiencewith with custom affinity(From product characteristics - i.e. what problems to solve)TheFood and beverage campaigns, visibility increased by 2 times.
（The table is a sample of the results of the analysis and the results compared to the industry average. showsSales growth in 10 perpendicular areas of the marketing industry by marketing goals and audience type. Example: Ad awareness increased by 2 times for audiences with similar interest in food and beverage customization, while ad recall rate increased by 1.3 times for audiences with similar interests in telecommunications. )
From the above results, we can find.In some industries, marketing targets have increased many campaigns, and the user base looks more clear.Recorded data also show that people who have bought computers in the market are inherently willing to buy.It is also normal for the indicator to rise high.
Relatively speaking, the following is more worthy of our attention:
Outdoor sports enthusiasts and fashionistas are the force of advertising recalls in telecom marketing campaignsBiggestOne of the enhanced affinity audiences.I'm afraid this is what many marketers in the telecommunications industry didn't expect.This also leads directly to missed opportunities to develop potential users.
Two lessons learned
The above conclusion does not emphasize the causal relationship, nor does it guarantee that it will achieve a bright result.
But two lessons learned from them are worth learning from:
The first is obvious:No longer rely on extensive data for marketing campaigns
Extensive user portraits, may be like a sieve, sift ingres sifting off your users;Granule user portraits(Segmentation, Segmentation, Subdivision)The smaller the graininess of the data, the greater the value of the commercial transformation may be released.
The second lesson is:Customer base probably much larger than you expected
Who would have thought that outdoor sports enthusiasts would love telecommunications advertising?Therefore, can not only rely on the surface, intuition perception of your users, the deeper the data, you from the user's real psychological is closer, the more likely to dig into more potential user groups, corresponding, conversion rate will also increase.
来源：《We analyzed 75,000 YouTube campaigns. Here’s what we learned about using demographic data》