Posted by & filed under PPC Chat.

This week’s PPCChat session was an interesting one. The topic was Facebook Ads and guest host Amy Bishop inquired about Favorite FB objective, Account structure to follow, segmentation of Ad sets and more.

 

Q1 – Do you have a favorite or go-to FB objective? Part. 1: If yes, which one and why? (Part 2 is coming for anyone who’s answer is no?

 

View Content is a strong objective if you have your pixel setup correctly. Great top of funnel objective. @zmste

Most of my clients either want (1) sales or (2) leads, so those are my go to favorites by default. @Ichasse

I find con works the best for our ecom clients but I always test traffic to make sure a client is not the exception. @duanebrown

runner up: App Installs. it’s not as good as other platforms, but it works. @JuliaVyse

Objectives can really vary based on what the client is trying to do… some that work well for Client A don’t for Client B and vice versa. But I do a lot with Traffic. @NeptuneMoon

conversions. I like website traffic too – I’ve also been using video views more recently for qualifying audiences. Just depends on what stage of the funnel we’re at! @DarthSamk

If I had to only run 1 I’d say conversion. You have a lot of flexibility to optimize for different events (Purchase, ATC, View Content, lead). Catalog sales, Lead forms, and event responses have their place but are much more circumstance or industry specific. @markpgus

I’m a conversions gal, so I’m into Store Visits and site conversions. BUT! I also like video views, then retarget the SV/Conv campaign to those who saw 50%+ of the video. @JuliaVyse

Not sure anyone has mentioned Reach, but I like to use it to keep my remarketing audiences warm so I’m not always hitting them with a lead ad. @SEMFlem

I usually get the best results from Conversion right now, but as we build out more TOFU content I’m looking forward to seeing what happens with Traffic as an objective teamed with a solid remarketing campaign. @HayleyDeee

 

Q2 – No surprise, it seems like conversion is the most frequently used but for those that don’t have a favorite or go-go FB objective, how do you decide which to use and when?

 

This is the part where I apply my listening skills because I contributed to the first question. @RyBen3

I typically build out conversion paths (some longer than others), depending on if education/nurturing is required. The objectives & content I use are dependent on the goal of the campaign & the client’s other resources (content assets, email programs, etc) @Hoffman8

I think traffic is a good test vs Conversion optimized to VC! If I have a conversion audience optimizing for purchase/value and see efficiency getting to VC or ATC then I’ll usually drop the conversion event and test. Next step would be to test traffic objective. @markpgus

I like to work back from Conversions. Where do I need ViewContent, Engagement, AddtoCarts, ect. This usually helps me understand where I can go with the budget that I have. @zmste

all the talk of video views! I use video views campaigns frequently bc it’s a great way to build really inexpensive a remarketing pool, & drive traffic at a low cost (often even less expensive than traffic campaigns). They can convert really well, too! @Hoffman8

Based on campaign objectives, but I like to test RF vs Auction or Traffic vs Conv where I can. @JuliaVyse

I based the metrics on (1) what my clients have set as a goal, (2) where we are start, mid, mature campaign, and (3) what other metrics support that goal (reach, views, etc…)  @Ichasse

If you’re not doing e-comm, then I think about what are we trying to make happen with the campaign and pick an objective that matches that most closely. Also depends on what’s available at that stage (FB pixel data, etc.). @NeptuneMoon

 

Q3 – When you set up accounts, do you have a particular account structure that you typically follow?

 

I segment based on prospect’s brand awareness: new-to-us / haven’t engaged; lightly engaged; heavily engaged; & have purchased. Depending upon the size of acct / amount of data I sometimes segment campaigns further but typically not until I have the performance data. @Hoffman8

Really depends on the market. I usually like to segment based on client objectives. (purchasers, engaged, brand building, etc…) @Ichasse

Love or hate CBO, it does work great for DPA campaigns I find. I like to break out DPA and IG Stories in their own campaign. Beyond that, I like to see purchase & prospecting based on how I plan to market stuff. @duanebrown

If budget allows, I try to remarket high-funnel campaigns w/ lower funnel direct-response objectives as “closers”. Saves money on the initial interaction and focuses the BOF budget on the most qualified traffic. @Hoffman8

Hard to answer bc each account is different. Half the accounts I start working with haven’t even set up pixels & have been running FB ads w/o them *mindblown*. Start small w/ optimizing pixels, then build out from there to acquire traffic, leads, conversions etc. Quick example. A brand I’m working with wasn’t excluding leads they had already acquired so they were wasting $2-3K per campaign. Simple exclusions have allowed us to acquire users at lower cost & reach new users w/o wasting spend on current users. @RyBen3

Basic Funnel template Cold Traffic: Interest LAL Non-Cookie Remarketing: Video View Engagement Cookie Remarketing: Site visitors VC ATC Past Purchasers. @markpgus

Don’t get complicated for sake of structure. Acquisition, Retargeting Campaigns then add more as needed. Each client/company is different, but Facebook is wanting smaller more concise campaigns. @zmste

Typically three tiers. 1. brand building/never engaged before. 2. somewhat engaged (75% video view, page visit, post interaction) – for example. 3. RM – navigated site, ATC, purchased previously…This can obviously get much more complicated than that though. @DarthSamk

I tend to view FB advertising as the new print advertising, so my thinking is often based on repeated messaging (though not the same ad) to build brand awareness & affinity. More salesy ads come later. Exception is custom audience from client. @NeptuneMoon

Cold audiences: Testing – interest targeting, smaller audience sizes, 10ish ad sets Scaling – combine the interests that worked in testing phase, larger audience sizes, 3ish ad sets. Bigger daily budgets & use CBO at this point And of course remarketing always! @HayleyDeee

TOF – Conversions TOF – VV PPE MOF (if applicable) BOF – Conversions/Cat Sales ADS LAB – Conversions (creative testing) ADS BANK – Conversion (build all original ads). @Gil_RunDMG

 

Q4 – Has the impending transition to CBO changed the way that you structure your campaigns?

 

Not yet. I’ve been using CBO more frequently in recent launches as preparation but I haven’t started creating single-ad-set-campaigns (SASCs?) by default yet. I have broken a few out after monitoring performance & noticing some weren’t getting their dues, though. @Hoffman8

Yes! now instead of one campaign with market level (=funding comes from) ad sets, now I have to do several campaigns per market. It’s so efficient!!!!! @JuliaVyse

I guess i figure if you had to make individual ad sets before then it’s not much different to make single ad set CBO campaigns… I’m not the one doing that work though so maybe it is. @markpgus

I LOVE naming conventions! I put the objective right in the names so I can differentiate between similar campaigns, ad sets, etc. @NeptuneMoon

CBO is a beast. I have learned that size of audiences plays a huge part in how much daily budget it gets. Overall I am finding more success with Ad Set Budget. Overall, less audiences, more campaigns. @zmste

Been starting to try to get in the habit of breaking out audiences by campaign where I previously would have been fine with them being in separate ad sets within the same campaign. @timothyjjensen

Totes. Try to keep campaigns dedicated to topic/targeting while making sure audiences sizes are similar…. now I have more campaigns but maybe that is not bad @duanebrown

No not yet. I just use the same approach as with Google ads. Split it up when you want to control your budget more specifically. @StephanieErne

Starting to put test audiences in a related CBO and then breakout the winners. The “always on” video prospecting audiences get their own campaigns. Otherwise, an up an comer can steal budget from the core audiences you always want to run. @jessesem

And when I have finally got my brain not fighting me on subcampaign budget levels in FB, I have to unlearn that thinking and go back to what should be the more familiar campaign level budget, except it isn’t quite like G Ads…@NeptuneMoon

So it did and then they pushed it out… However, now I always start testing with CBO and will you use spend limits if I need to force traffic. Throw 3 interest categories in a CBO find a winner and then test Ad set budgets. @markpgus

Not yet, apart from me trying it every few weeks to see if it will work in an account. For scaling, one new thing that can work if an adset has done well in ABO is duping it 4/5X into a CBO at around $300/d and letting it ride for 3 days before doing anything with it. @Gil_RunDMG

It has made me make smaller more focused campaigns to see results quicker and then expand the budget in the next campaign. I try to think of it as quick a/b testing. @RyBen3

At the moment I’m only putting winning interests into CBO campaigns… that’ll change when CBO is my only option! I’ve also seen that fewer adsets with larger audience sizes in them tend to do better. Too many adsets in one CBO campaign seems to throw things off. @HayleyDeee

We’ve been completely CBO for some time. We trust Facebook to understand the best way to distribute budget and optimize performance and it’s been plain sailing ever since we did. Our mantra is to keep it simple and CBO = simple. @jimbanks

 

Q5 – When you structure your ad sets, how segmented do you typically start with your ad ad sets? Do you segment ad sets by platform, device and/or placement by default?

 

Usually it’s campaign objective, then audience name in ad set name. Sometimes city/region if needed. def not by device. that’s too much. @JuliaVyse

Generally thinking about segmentation by audience categories. I tend to wait to break out by device/platform/placement until the data justifies it. Depends on the campaign, though (sometimes go in more granular up front based on prior data, creative, etc.) @timothyjjensen

I generally don’t go too granular (platform, device, etc.) at the beginning. @NeptuneMoon

Different across clients b/c some know their audience, others don’t. For those that know, I target by broad interests but really nail the demos of that buyer persona. ie. Male/Female, Age range, interests. Haven’t had large enough ad spend to segment more than that. @RyBen3

With the new Facebook mobile newsfeed ad size. Breaking out by channel (fb vs ig) may make even less sense starting next week. @duanebrown

Campaign number codings, 3-2-1-0, for different parts of the funnel and goal. Ad set naming by ad type, audience, placement, and location. IG stories always get their own ad set. Splitting out IG feed from FB feed depends on the ad unit. @jessesem

Typically break out by objective & interest/targeting. Not really by device/platform – I could see that changing though if we notice something outperforming though. @DarthSamK

If the acct is big enough to where there will be enough data, I like to segment by device / platform / placement but if there isn’t enough data, then I keep things combined to give FB enough data to make decisions. @Hoffman8

I’m moving away from segmentation and am bought in on account simplification… Lowers my required bandwidth, lower cpms, With the 50 conv/week recommendation I’m getting a lot broader…. This is admittidely tougher for SMBs. @markpgus

We like to segment by age/gender even though it’s splitting a bigger audience. We find that age/gender can impact significantly performance and we can play with ads and talk better to audiences that are 25-34 F vs. 25+ All Genders. @jimbanks

It’s absolutely worth breaking stories out, imo, as they perform much differently & setting up different story-worthy creative is imperative. @Hoffman8

I usually start with all, then segment based on performance. For certain brands, separating M/F out and using more personalized messaging makes a lot of sense. Always good to test Stories on their own too. @Gil_RunDMG

No we usually start with not segmenting those things. We do that later or depending on how long the campaigns is we take the learnings for a next similar campaign @StephanieErne

 

Q6 – What’s your favorite way to identify interest targets?

 

Lookalikes of high value audiences x 10000! But in addition to that, building out personas & using the audience insights tool to help flesh them out is a great way to identify targets. @Hoffman8

I’ll still V-scale the ABO by up to 20% a day but be more aggressive with the CBO. If I duped the original 5X, after 2-3 days there’s usually one or two obvious candidates to kill which leaves around 3 good adsets to run. @Gil_RunDMG

Audience Insights! Even though Facebook has taken away some of its functionality I still use it regularly to help me build out audiences identifying – age, gender, marital status, professions, locations. It is a gold mine when you are not sure where to start. @zmste

Lookalikes are hands down the most powerful thing FB does. In all marketing we always want to find more people that are are “just like our good customers”. FB has the ability to actually do it (if you have a good source list). @NeptuneMoon

I’m also a fan of Facebook’s suggestions for targeting. It’s a great way to branch into new interests that you might not have thought of. @Hoffman8

Be market specific but using FB/IG insights too. If you are advertising someone in NYC that’s interested in baseball, you probably can’t go by just baseball. You have the Yankees, Mets, Blue Jays, Red Sox fans all in the same area. Need to get more specific. @RyBen3

Audience Insights. @HayleyDeee

Apart from the tons of different LLA you can create; digging round audience insights for half an hour can get you some useful data. I also get suggestions from the client as part of the onboarding process. Key is to always validate by the data! @Gil_RunDMG

Use a tool like Connectio or the Facebook Graph API @jimbanks

1. Audience Insights on page likes. 2. GA data 3. Client Brain Dump 4. Look at post engager profiles and get an idea of who they are and if there are trends… this is very unscientific I know. Wish I had Cambridge analytica scraping ability. @markpgus

I tend to lead with common sense generated by targetgroup insights the client has themselves first. Later and if available throw in some data we got from Facebook, linkedIn, Google or other platforms that give some audience insights. @StephanieErne

1) The insights I get from the brand Instagram accounts through FB analytics are 2) What else are our influencers into and what brands are hitting them up for deals? 3) Your external sales teams are an amazing asset for learning about your customers.@jessesem

 

Q7 – What’s a strategy you tried or test you ran (could be anything – bidding, creative, structure, targeting, etc) that was either wildly successful OR that was a big learning opportunity?

 

A strategy that worked really well, that I use frequently now, was to set up an engagement campaign using a custom audience to target ONLY happy customers to drive a bunch of engagement/comments & then using that ad ID in prospecting campaigns. @Hoffman8

Had a client who had audiences kept failing. I pulled a list of all of their orders and used a heat map program online that plotted them for me. From that I created custom audiences for just those locations per state and had amazing success for about 3 months. @zmste

I took audience insights from YouTube (what ppl watched when my ads ran) and then flipped those into FB interests. #humblebrag I included this in my heroconf deck because it WORKED!!! @JuliaVyse

Working with a client who did not have a Black Friday/Cyber Monday offer, we switched everything off and let the crazies beat the crap out of each other. Client thanked us, paid us a healthy bonus. @jimbanks

Pinpoint very specific area’s where you know people that are visiting the area are top of funnel. Example: People visiting the area where their new house is being build as part of a house development project. For advertising for gardeners or home decorating services. @StephanieErne

With a Canadian ecom client. The huge difference in audience performance with customer emails vs purchase pixel. My mind says they should perform similar. Facebook has the latter shatter the former like nothing I have seen. @duanebrown

IF YOUR CLIENT DOES EVENTS GET IN ON THAT!!!! Help them short term and build audiences long-term. Event responses retargeting to order page FTW. @markpgus

 

Q8 – If you could give your top 1 or 2 best tips on Facebook Ads, what would they be?

 

Test everything, all the time. Pay @Hoffman8 to manage your accounts. @AndrewPPC

Try not to over-segment. If ad sets are small & don’t drive many conv, try optimizing toward a micro-conv in order to fuel your bidding strategy w/ enough data & monitor the impact on lower-funnel conv & remarket. @Hoffman8

Be as creative and inventive as you can in targeting and content! This creates the most fun and quality campaigns. @StephaineErne

If you have something segmented, you better have specific creative to complement it. Creative is probably the largest lever we have. Segmenting audiences without tailoring creative is half assing it @markpgus

The key to learning from FB ads is testing. EVERY DAMN DAY. The second is don’t think you know everything. Constantly ask people for help or participate in these chats. You learn so much from other people and it gives you ideas on what you should be testing. @RyBen3

Don’t underestimate how much creative you need ready to go when you start running campaigns! @NeptuneMoon

Get the pixel installed on your site and track any core actions that you can identify @timothyjjensen

Find your own “best practices” and understand they may vary by account. Learn from others and don’t be afraid to ask others for help/insights (we can save you pain and frustration if we’ve been there already!). @NeptuneMoon

Don’t forget to test different objectives & get creative with how you use them, you might be surprised. @Hoffman8

Use standard and custom events to track every significant action on your website / lp to fuel the pixel. @mickymereu

Spend time in the Audience section, build out Saved, Custom, Lookalikes. Facebook is constantly changing and offering new things. 2nd, learn and use Facebook rules as you can set them up as notification emails in the morning. @zmste

 

PPCChat Participants:

 

Related Links:

 

 

Leave a Reply

Your email address will not be published. Required fields are marked *

Website Protected by Spam Master