AI making moves to dominate PPC advertising

Despite the tremendous demand for PPC advertising talent, there are already signs that smart machines are moving in to take over this lucrative business. The Drum published an article titled “IBM starts using Watson AI to buy online media in the UK” that provides a preview of a machine-driven future for PPC, SEM and other paid marketing activities:

Sitting within The Trade Desk – its media planning platform – Watson over time learns how effectively a campaign is performing for different audiences at different times, locations, devices and browser.

Based on this information, it will then only bid on inventory that aligns to any given audience, and even then will consider the size of an ad and how effective it will be in relation to those other factors.

In the US, IBM claims this has reduced its cost per click by as much as 71%, although the average hovers around the 31% mark.

This should give performance marketers something to think about. I doubt the demand for SEM or PPC advertising professionals will diminish in the very near future but I can definitely see a time when smart machines like Watson will take over paid media and put a lot of people out of work.

Postscript: Take a look at how Watson is starting to impact influencer marketing too in the Drum’s article titled “The Movie Marketing Blog: IBM Watson-powered Hollywood influencers“.

Image credit: Robot Mural from MMT

The trouble with LinkedIn engagement

Yam Regev shared some interesting data about LinkedIn engagement on LinkedIn.

My LinkedIn activity stats (on average):
– Textual post: 3k views
– Post with a video: 4K views
– Post with an image: 6.5k views

  • Published article: 80 views

You read it right, it is generating only 80 views.

If I put so much effort in it, why do I get only 1% of the views I see from other kind of posts?

To have more traction and content on its platform, LinkedIn will have to change this realty.

I tried to share my thoughts about LinkedIn engagement with him in the comments but it seems my feedback was too long for the comment box so I shared a shortened version instead. Here is my original comment:

I don’t think this is limited to LinkedIn. The major social networks optimize the interactions that keep people on the sites and clicking. I suspect text posts, videos and image posts keep people moving through the site. They are fairly low friction content pieces, easy to engage with through likes and comments and easy to consume before moving on to more. They’re the digital equivalent of snacking – high volume.

When it comes to articles, they are longer form and people need to dedicate more time to reading the articles. It would be interesting to have % complete metrics like you have on Medium to see how much of an article people read before abandoning it.

You’d probably find that people view longer articles less frequently because they prefer to snack on social media and not spend time reading the articles in their entirety. The problem with this is that it tends to discourage longer form content. It also doesn’t help all that much to write longer form stuff elsewhere and share it here because these services optimize for keeping people in the service, not making it easier for them to leave for sites outside the social network.

Image credit: kaboompics