September 2017 / Page 0marketing.scienceconsulting group, inc.
linkedin.com/in/augustinefou
Pricepoints and Network Study
KEY FINDINGS
1. The levels of bots and humans did not vary by price point (programmatic CPM); these
depend more on the quality of the network (not shown: 1¢ CPMs had very high bots)
2. The highest CPM has more humans, but it also has more bots in Network 2
3. There was NO difference in viewability across price points; it depends on the network
NETWORK 1 NETWORK 2
$0.50
$1.00
$5.00
$10.00
Humans (dark blue) vs Bots (dark red) by CPM
58% average 45% average
Viewability (desktop vs mobile)
Marketing Science studied impressions at 4 different CPM price points, across two ad
networks to see if there were differences in the levels of bots versus humans. There was
no significant difference; the quality depended on the network. But definitely stay away
from super low cost inventory – i.e. less than 50¢. It’s mostly fraud and bots, as expected.

Sept 2017 - Ad Fraud Pricepoints Study

  • 1.
    September 2017 /Page 0marketing.scienceconsulting group, inc. linkedin.com/in/augustinefou Pricepoints and Network Study KEY FINDINGS 1. The levels of bots and humans did not vary by price point (programmatic CPM); these depend more on the quality of the network (not shown: 1¢ CPMs had very high bots) 2. The highest CPM has more humans, but it also has more bots in Network 2 3. There was NO difference in viewability across price points; it depends on the network NETWORK 1 NETWORK 2 $0.50 $1.00 $5.00 $10.00 Humans (dark blue) vs Bots (dark red) by CPM 58% average 45% average Viewability (desktop vs mobile) Marketing Science studied impressions at 4 different CPM price points, across two ad networks to see if there were differences in the levels of bots versus humans. There was no significant difference; the quality depended on the network. But definitely stay away from super low cost inventory – i.e. less than 50¢. It’s mostly fraud and bots, as expected.