Conversion Rate Drops - What Do You Do? (6)
I know this could be dependant on the traffic source, geo, the offer itself etc.
But lets talk on a fundamental level, assuming nothing changes on the traffic source, the geo hungry and the offer is evergreen.
You have an offer thats converting well, get it up to $XXXX/day in one geo, but then things slow down, other geos are ok — not great.
What do you test first to revive or keep the campaign going?
I'm testing new banner images, headlines, landing pages—it's a scattered approach, so I'm interested to hear from more experienced marketers what they do when they see an offer C.R. dropping.
Not at all as experienced in AM as the rest of the folks here, but from what I've read here and also from personal testing, it really depends
on the questions you ask yourself about that campaign:
1. have you done anything prior to this drop to the campaign? bids, LPs, server stuff, budget wise, placement cutting?
2. is the type of offer something that is evergreen?
3. is it worth reviving it vs finding a new one using the same process you did when you first found this one? here I'd say try the later
4. has this offer gotten more competition?
From what I see you're using display and prelanders so even more variables come into the mix. Ad blindness comes to mind ( which is more common than you think ).
Here you'd be better of buying yourself Cashvertising or Words that Sell. Those guys have the right advice on this. Little changes to your banners ( colours, placement etc )
can drastically reduce ad blindness and actually improve CR ( further testing is however necessary )
Daniel
testing a new banner images, headlines and landing pages should be the way, but what about testing a new traffic sources?
First you determine what was done when conversion rate dropped, did you change something (ie budget), did the CPMs go up, did the CTR drop?
Finding out the cause, can help pinpoint the problem.
If the CPMS/CPC went up, maybe you increased too fast the budget. Or maybe a new competitor came in with a better performing banner set.
If the CTR dropped, then people are getting bored of your banners, so the algo is showing it less. Time to refresh creatives.
One tip, if you can graph on chart so you can see it visually, things like CPM, CTR, CPC, etc you might be able to spot what happened/where, at what life cycle, to be able to come up with a solution/plan of attack.
This is not so easy to "diagnose" because you only know about changes you have done yourself. What happens at the traffic source level remains a mysery very often.
There are several possible reason for this :
1. banner/creative burnout, this is the way I understand it : http://stmforum.com/forum/showthread...-down-and-down
This problem is the easiest to overcome, because you just need to find a new set of good performing creatives ... I know it can take time, butth solution is in your hands so to speak and yo control it to the full extent.
2. change on the offer side - even small change in the funnel, new regulation, new requirements on retention ... you need to ask the AM if they know about something. In case they changed something, there is not much you can do, other than tweaking the funnel and trying to come up with a new approach. In many cases, this is game over and you need to find a new offer.
3. change at the traffic source - this depends on the vertical and the type of source. Adult is an excellent example for this so let's use it - there are mega sites that buy traffic, have returning visitors and get a lot of organic search traffic. In case google makes a change to the algo, they can loose significant amount of the search traffic but in order to keep the same numbers, they start buying more traffic to compensate for the loss. Obviously, the traffic quality went down now and it will convert worse. Again, nothing you can do here, just optimize better.
4. competition - all it takes for someone to copy your funnel entirely and bid more than you. Or even match your bid with a different funnel, this can affect your CVR because the traffic distribution changes. Go down a level or two in the bidding chain and you will loose quality. Large drops are always connected with traffic drop too, but with some networks, the drops are not so significant when you only go down by 1 level.
Traffic sources are dynamic beasts, the conditions are not stale, they change all the time. Algos are tweaked all the time, placements come and go, bids go up and down ... all of this can have some impact on your campaigns.
Of course there are also the changes you have done yourself, guys above me mentioned most of them so no need to repeat them. Just make sure you are aware of every single change you have done, even the small ones like upping budget, changing distribution model (throughout the day to accelerated), capping ... or even factors like how much your balance is and what kinda daily spend caps you are using.
To sum it up, these drops are usually caused by one of the 2 main factors : 1. Burnout and 2. The way ads are distributed. #1 is fully in your hands, #2 only partially by changing the campaign setup, the rest is on the networks shoulders 
Interesting points.
Yeah I read that post on Banner Burnout last week—great post Matuloo! I had the same frustration for a while, but switched out banners and CR increased again.
I did a quick analysis and CTR has been on a downward trend, so lets see if new banners will help.
I added a new lander test + I'll write up a new pre-sell.
Im in the fortunate position to be running exclusive and direct so I've got a handle on the offer changes and this pretty much eliminates the competition element for that offer.
And I'm also testing a new traffic source.
Home > General > Affiliate Marketing Forum