A lot is discussed about personalization, but how accurate is the information being used to market to customers individually?
I recently jumped on the Oracle bluekai registry last week to find out. While there were some things that were correct about who I am, there are many things that were completely wrong.
- I'm not a baby boomer
- I'm not a huge car enthusiast
- Not looking to go on a diet or lose weight
- Not in the market for jewelry or watches right now
- I have not bought gardening or landscaping items recently
- I'm not really into hiking and camping
If you are a guy who is married with no kids, you may wonder why you are being marketed baby onesies over the Internet. You may ask yourself one of these three questions:
a) Was my wife on here earlier and maybe she is thinking about starting a family sooner than I thought?
b) Wasn't my sister last night looking for a baby-shower gift for her best friend, Amy?
c) Does this have anything to do with liking an ad my buddy's newborn was in last week for OshKosh?
It could be all or none of these actions on your computer have occurred and were tracked. The point here is that there are multiple reasons why you go to various websites, and there could be multiple users on our electronic devices that interact with the internet differently than we do ourselves. This is why there is so much misinformation on each of us demographically.
This is why first and second-party data need to be leveraged more by ecommerce sites to get a better return on advertise spending (ROAS).
The problem with third-party data is that there is definitely some useful information for a marketing professional to extrapolate from, but there is also, unfortunately, many questions you need to ask a third-party data provider before you commit to the partnership:
1. How do they collect data? Is this a cookie-specific collecting, or do they go beyond the cookie-centric data gathering approach and have insights into a particular user's offline activity? What are the factors, actions and metrics determine whether a user falls into one segment or another?
2. What is the time frame that a user needs to have taken such actions for him/her to be considered a viable candidate for the company's third-party data collection and documentation?
3. How current is the data? What is the time span that the provider updates their cookie pools? Next, inquire what is the schedule for syncing up their cookie list to its DSP (Demand Side Platform)?
4. How are these data segments compiled? Multiple sources or a single one? Using various sources can lead to a vast range in the quality of this data being collected versus data from one specific source will be a lot more consistent, generally.
And even when you get a reputable third-party data provider, many times the data you are getting access to is at best an educated guess. Sure, if a user jumps on a publisher's site and reads an article about a luxury vehicle, this user may have an interest in buying a luxury car soon or may simply like luxury automobiles in general.
What third-party providers do have going for themselves is the monumental amount of data they collect. But that's both a blessing and a curse for marketers who are seeking out this data. While getting data on millions upon millions of users sounds enticing, will making the financial commitment to attain this data be worthwhile for a marketer's goals?
It depends, really. Is the marketing campaign being put together, based off of this third-party data, having to do with top-of-funnel or more bottom-of-funnel goals? Third-party data can be used for either of these types of marketing campaigns, but a marketer needs to make certain that he/she has a clear vision and understanding of how to leverage this partner's data, correctly and have reasonable expectations.
While you can definitely reach a lot more people using third-party data over using first or second-party data, you will also be reaching a lot of users who do not care about your service or product. Why? Because third-party data is based off demographics and not off of a user's intent.
Conversely, first-party and second-party data will not give you access to as many users as third-party data, and do not care as much about a user from a demographic standpoint, but more on an intent perspective.
No. Third-party data doesn't really know who you are. Inferences can be made. Logic may be applied and assumptions are most definitely occurring. But, I know that the internet sees me (on my computer at work) as a 50+ to 60-year-old man, who enjoys cars, needs to lose weight, likes gardening and is planning a camping trip, perhaps...
No, that is not me.
Bio: David Allison is the "Marketing Guy" at Hivewyre, which is an ecommerce data cooperative in Scottsdale, Arizona that deals solely in second-party data.