Facebook Suggested Bid Spiral Effect Part II

by redembermarketing on July 10th, 2009

DON’T CHASE YOUR TAIL
So I got my response from Facebook, and there are some good and some bad parts. The good news is I have a new strategy to try, the bad news is that not only is it extra work, there are implications for people that have different offers targeting the same audience. Let’s start with Facebook’s response:

“I’ve checked your ads and they seem to be functioning well. However, I’ve also noticed that your ads are targeting the same target audience. Please note that for any given ad unit, our ad auction system selects the best ad to run based on the ads’ maximum bids and ad performance. If you have two or more ads in your account, targeting the same audience they will compete against each other in the auction. If one of the ads performs better than the other, it will have more chances of winning the auction while the other ad may get less impressions as a result. Since the bid is influenced by the performance of the ad, this may increase the suggested bid for this ad.”

So I was chasing my own tail? I hadn’t thought of my ads as “competing with each other,” particularly since they are in the same campaign, and like most marketers, I’d like to be able to test the performance of more than one ad, preferably against the same audience. Apparently, that’s not really possible unless you run them exclusively, at different times. Suggestion to Facebook: let advertisers display ads evenly to expedite the process that identifies the best performing ads. Who wants to compete with themselves? I hope this makes its way up the development track for fixing.

Well, I did say there was good news, and it’s that there’s a new approach to try. It’s actually Facebook’s idea:

“We recommend separating your ads to target different states. You can then create ads that are more targeted to each of your audiences. This will ensure that your ads are not only not competing with each other but also that they are more relevant and accurate.”

Since my ads target specific states, but the same demographics in each state, so I guess I could create state-specific ads, but seriously, up to 50 different ads just so they don’t compete with each other?  Yeah, they’ll be more relevant, but perhaps not enough to justify the extra effort of managing so many more ads (and aggregating the data from each for reporting).

Facebook Suggested Bid Spiral Effect

by redembermarketing on July 9th, 2009

I was debating whether or not to blog about this, but since a few people have confirmed the phenomenon, and I couldn’t find any other blog posts about it, perhaps it will make an interesting story.

Background:
I run some Facebook Ad Campaigns for clients, and have recently noticed that impressions were drying up.   At the beginning of the month, I had created some brand new ads and increased bids slightly, so I didn’t check the suggested bid range before contacting Facebook support.  Their response was that the ads were not being displayed because bids were not in the suggested range.  Of course, they were at the high end of the range when the ads were created, but within just a few days, the range changed fairly drastically.

So I increased the bids a second time and checked back the next day.  Again, almost no impressions – I was surprised to see that the range had increased again!  At this point, I start to get suspicious, and started contacting some other PPC colleagues to see if anyone else was experiencing this.  There were mixed responses, but at least one person confirmed what I am now calling the suggested bid spiral effect:

“I am seeing this spiral effect too. However, unlike with Google’s AdWords, it is impossible to trace it to a specific driver. As you note, Facebook’s lack of transparency is frustrating to say the least. They make changes without telling anyone and come across as totally capricious. I actually wonder if they are in trouble because it is as if some executive is cracking the whip in the background fanatically trying to extract more and more money out of each interaction with zero regard for stakeholders. Until they explain the rules better, Facebook ads will remain an unviable option for serious advertisers IMO.”

With little change in impressions and no change in the CTR, it’s tough to figure out what’s driving the bid range increases.  I decided to continue playing the game though, only this time I’d monitor it more closely.  No sooner than a couple of hours after raising bids, I noticed that the range increased again for some of the ads.

I should have noted this earlier, but all of the ads were set up as CPC.  So my next tactic was to create a similar ad and change the bid type to CPM and see if it made a difference.  Over the next 24 hours, the CPM Ad represented 75% of the clicks among 5 other ads.

Although I have heard nothing back from Facebook support about the issue yet, perhaps others will confirm the same behavior or describe what they did in response.  The mystery surrounding this simply calls attention to the obvious lack of transparency and documentation from Facebook about how they run ad campaigns.  The more information/guidance they can provide to advertisers, the more confidence the market will have in their platform, not to mention less negative commentary like this blog post.

4 Tips to Maximize Your PPC Campaign Investment, Post-Click

by redembermarketing on June 3rd, 2009

So you’ve been refining and adjusting your PPC campaigns for a while now.  Cost per clicks are coming down, bid management is under control, you have strong offers and optimized landing pages.  Conversions are up, and leads are coming in.  Sounds like lead gen uptopia, right?

Well, almost.  PPC campaigns are feeding information into the very earliest stage of the sales funnel.  What are you doing with that information to work newly discovered leads down the funnel, identify the hot prospects and close more deals?

Here are 4 tips for maintaining the momentum of the lead gen and sales process once someone has clicked on your ad:

CREATE A SEGMENTATION “LANDING EXPERIENCE”
You’re probably familiar with landing pages, and perhaps you’re even experimenting with optimizing them to maximize conversions.  But do you ever get the sense that not all of your traffic is exactly what you were aiming for?  Perhaps you’re products or services span several verticals or customer demographics and there’s some overlap in your campaign generated traffic.

To help make sure the right visitors get to the right pages and see the intended content, consider directing traffic to a segmentation page that makes it easy for visitors to self-identify themselves.  Two large images with short simple text can help people identify which group they belong in, at which point, they are served the correct information.

This is the start of a “landing experience” – a microsite if you will that visitors interact with before submitting their information or making a purchase.  This can actually help increase conversions since your leads are served more appropriate, targeted information.

SAY THANK YOU, PERSONALLY
Are your leads or new customers being taken care of once they’ve converted on your site?  Conversion might mean submitting a form, watching a video, or making a purchase, but are you taking them to a thank you page after conversion that introduces them to the next step in the conversion funnel?  The landing experience doesn’t have to end once someone has converted on your site.  Put some additional thought into how you say thank you and increase return traffic and repeat customers.

You could provide them with a special offer related to the offer they just took advantage of or the purchase they just made.  Of course, the thank you / follow up email is also a great opportunity to provide the customer with well-targeted offers since you now know more about them.  Test the timing of follow up emails to see when the optimal time is to send those messages.

DEVELOP A PROSPECT PROFILE
Depending on your budget, it may be harder to develop a well defined prospect profile without special software or technology installed on your website to track visitor behavior, notify sales when specific pages are viewed, etc.  But you can still come up with ways to learn more about your prospects.  They’re in your database, so you should know something about them (especially if you’re list is opt-in, right?).  The important thing to identify is what you don’t know about them and come up with ways to fill in the blanks.

Surveys are great for this.  You can develop a customer satisfaction survey, preferences survey, a contest, etc. that blends important and not-so-important questions together so you can extract useful information.  Let’s say you’re a florist – you can have prospects vote on a series of arrangements that will be offered in the store.  The results from the participants are invaluable to you!

Or, perhaps you’re B2B -  a software company – you can create a survey that not only collects information about your prospects so you can market to them better, but the results could be turned into an industry report and picked up by the media.  SEO score!  [Consider having a credible third party assist you with collecting the data / creating the report so it's a bit more objective]

NURTURE LEADS APPROPRIATELY
For some prospects, your special offer on the thank you page won’t prompt additional action, and in many cases your thank you / welcome emails may appear to be falling on deaf ears, but don’t stop nurturing.  No, I didn’t say keep spamming them, nurture them.  Change up the offers that are being presented until something prompts a response (preferably a purchase or a call, not a Spam Report).

By timing nurturing campaigns to spread messages out over days/weeks, you’re less likely to annoy someone and catch them in a different mood/business position.  This takes prep work ahead of time, but it’s worth it when you know that inactive leads or dormant customers are receiving targeted content that you crafted and should work…  Eventually.