Mumbrella360 video: Determining the true ROI for your media decisions
In the latest in a series to run over the next few months, we present highlights from this year’s Mumbrella360 conference.
Andrew Palmer (Executive Director, Nielsen) and Mary Ann Azer (Executive Director, Magazines Publishers Australia)
Knowing the true ROI of your media activities could have a huge impact on what decisions you make. In this Mumbrella360 session Andrew Palmer, executive director at Nielsen, and Mary Ann Azer, executive director, Magazines Publishers Australia, share their local global learnings in the area of econometric modelling.
Timeline:
- 0.44 – Why are we here today?
- 1.20 – Not just where we spend, but how we spend
- 2.10 – What is econometric modelling?
- 2.51 – What output do we get? Can we predict sales?
- 4.05 – “Let’s talk about ROI and the bigger picture”. The five key questions.
- 6.09 – “Data is the key MPA story”
- 8.14 – “The data input matters”
- 8.41 – “I’ll explain what MAPP is”
- 10.23 – Put the data into models to see if it is working or not
- 11.20 – “We’re building two sets of models”
- 12.00 – Media contribution to sales revenue
- 13.50 – “What does that mean in terms of ROI?”
- 14.30 – “Makes a very different investment decision”
- 15.10 – Simple synergies: 1+1=3
- 16.14 – “The difference is now quite important in terms of how I lay down my media”
- 17.05 – “Misattribution and how we think about it in modelling”
- 19.09 – “We need to think about the network effect”
- 20.42 – Long term effects and the sticky nature of the media
- 21.27 – “How do I value my media?”
- 23.06 – There is considerable variably between brands even within an industry vertical
- 24.02 – “Copy is King. It does affect your ROIs”
- 25.08 – “We see a direct correlation between stronger creative and actual ROIs”
- 26.10 – “Effectiveness is not the same as ROI”
- 26.52 – “This is where modelling will go. We’re going to start building engagement indices”
- 27.47 – The difference between reach and reaction
- 28.37 – “A reach curve is not a response curve”
- 31.14 – “Focus less on reach and circulation and more on response”
- 32.20 – “How do you manage these recommendations and the relationships with the media agencies?”
- 33.42 – “What do you mean by ‘quality copy’?”
- 35.15 – “Do you factor macro economic effects into the models?”
- 36.55 – “How do you measure brand equity?”
Oh give me a freakin break. The fact that they call it “media roi” reeks of bias. Bias towards media. What if those media investments have no roi, will Nielsen still report on that or will they manipulate the data to create a roi because that’s what their client paid for?
Another problem with these models is that you cannot control for all the independent variables that has a relationship with sales because a lot of these independent variables are unquantifiable, as such what you will get is a magnified effect for media variables to compensate for the missing independent variables in the model and they do this with lags and carry over……and in some cases dummy variables because they have to model a ‘media roi’ by all means possible.
As such, if you are doing econometric modelling, make sure you ask them to explain the carry over and lag and determine for yourself if it is justifiable. Also ask them to give you a list of all the variables that goes into the model , including the dummy variables.
Also ask them if they do proper data screening and assumptions testing, which I doubt they do.
If this econometric bs works then surely economist will be able to predict the economy ………….
User ID not verified.