Step 1: Selective Bidding Menu
View By Day:
Sort by total clicks sent:
Look at the top spending Ids.
Are any of the top 10 Ids spending 10% or more or the total traffic?
Down bid them .1 – .2 to slow them down a bit, but not lose quality.
Adjust the bid based on the volume/performance and most importantly. Relatively compared to the rest of the sources.
If the top spending ID is spending $200 a day.
The second Id is spending $175 per day, and the 3rd is spending $70 a day.
Drop id 1 and 2 from 1.0 to 0.8
Drop id 3 from 1.0 to 0.9
The general idea is to get an even flow of traffic hitting as many unique ids as you can. Good ids tend convert and show themselves very early in testing.
Sort by Conversions:
Adjusting bids here will depend on the performance of the source, amount of daily clicks, and the margins you have.
- If one id drove 50 clicks, 1 lead – over 7 days. With a 40% ROI. You could safely adjust that bid up from 1.0 to 1.2 or 1.3. It’s not spending super fast, its converting, and you want to try to get more out of it.
- If one ID drove 150 clicks, 1 lead, over a few days. With 10% ROI, you would want to down bid that one from 1.0 to 0.9
This is because you don’t have the profit margins to pay more for clicks and be profitable. Slowing it down, and spending less may make it greener. Ads could be the bottleneck as well.
Important: Any new placement that converts quickly, is a gem worth spending to test more. Give them a small bid adjustment, and keep focusing the system into high performing placements.
Sort by Profit Loss:
If you optimized based on daily performance on profit loss. You can end up turning off tons of profitable ids. If you do this, be sure to review “all” data. Sort by conversions. Make sure you didn’t turn off anyone that’s really decent.
- Normally you will want to optimize this on at least 7 days worth of data.
- If you do not have proven ads – let them spend at least 85% of the CPA.
- If you have proven ads – let them spend at least 70% of the CPA
- If you have some ids spending Very heavily that day and not performing well. Pause them.
- If a widget is converting but not profitable: Look at the avg cost per click for that unique widget, and the EPCs its generating. Adjust the CPC closer to where the break-even point is.
Step 2: Larger Time frame
Repeat the same steps above for “yesterdays” report, last 7 days, and all data. The only unique thing you do here, is look at the last edit date/amount adjusted of the widget. If you have a converting widget that has received a bid adjustment, and if it’s not gaining more traction. If it is been a few days, give it another one.
If profitable widgets are not spending any more, it very well could be from the competition. Review these and give them adjustments. Nobody does this, and its how you really can scale up easily. Just from gaining back all your old positions.
Step 3: Ads
Spend: Has the ad spent the total cost of what one conversion pays? Do you have widgets semi-optimized? If so, then pause it.
CTR: Normally you are wanting to drive above .150 ctr on most ads. If the ad is way below that after testing, (under .04) and not performing. It is super unlikely that it will turn around. Kill it, its just wasting impressions.
Reach: Incredibly helpful. When you start seeing the weight of the ad grow, and it hit low/med reach. You will see a major difference in quality from very low. Very low is only accessing under 30% of the total inventory for your unique campaign.
If you have a good ad and its performing for you, this will let you know how well it is doing in the system. How many placements it has access to.
What that means for you: If your ads are all very low reach, and can’t get into low/medium reach. Could be bidding too low, or more likely. Just not the best images and headlines. Try to make them more enticing.
If you have an ad with low reach, that’s breaking even and has been stable. It could and likely will turn around once it gets onto higher quality placements.
If all your newer ads are “very high” after the first blast of traffic, plus another 12-24 hours. It’s possible that you are overbidding.
Avg Cost per Conversion: This will let you know how stable this ad has been performing on the current widget optimization. Ads that are stable and perform don’t just magically happen. Even if the cost is higher than you want, look into what’s unique about this ad. If the daily costs are getting lower, then your traffic optimization edits are working.
Today’s Cost per conversion: The traffic you get today generally depends on what’s been done during the last few days. This is why you need to take things slow, and not brute force it.