{"id":7211,"date":"2018-10-26T13:27:29","date_gmt":"2018-10-26T07:57:29","guid":{"rendered":"https:\/\/www.karooya.com\/blog\/?p=7211"},"modified":"2018-10-26T13:27:29","modified_gmt":"2018-10-26T07:57:29","slug":"ppcchat-attribution-oct-23-2018","status":"publish","type":"post","link":"https:\/\/www.karooya.com\/blog\/ppcchat-attribution-oct-23-2018\/","title":{"rendered":"PPCChat | Attribution | Oct 23, 2018"},"content":{"rendered":"

This week’s PPCChat session was hosted by Elevated Marketing<\/a>.\u00a0PPC experts discussed about attribution, how they used this metric, which attribution model they prefer in Paid Media, multi-channel attribution challenges, how it affects budgeting and more.<\/p>\n

 <\/p>\n

Q1 Since there are so many variations on what attribution means\u2026… let’s start with how would you describe the term \u201cattribution\u201d as it relates to paid ads?<\/strong><\/h4>\n

 <\/p>\n

For us It is the tracking on attributing success (conversions) to either a keyword or ad or on a broader scale attributing success to different types of paid ads channels ( Facebook, Google ads, Twitter etc.)\u00a0We change all of our to 28 day click and view to get the most credit for our ads\u00a0 \u2013\u00a0@elevatedmrktng<\/span><\/a><\/p>\n

The really short answer: To which stage in the customer journey you attribute the conversion. Usually I explain it with examples of different conversion paths.\u00a0\u2013\u00a0@StephanieErne<\/span><\/a><\/p>\n

Pretty simple definition. The effect an effort had on an end result.\u00a0Attribution is a really important topic. HOWEVER, there are no correct answers. We don\u2019t have a perfect tool. What is important is that a method is chosen and it is well thought through with the understanding it is far from perfect \u2013\u00a0@markpgus<\/span><\/a><\/p>\n

It’s a way to figure out the answer to the question of “which half of my marketing budget is wasted.”\u00a0\u2013\u00a0@ferkungamaboobo<\/span><\/a><\/p>\n

It’s critical to have an understanding of attribution across your channels to run effective and scalable paid campaigns in 2018. Knowing how your channels correlate with each other is often the difference between success and failure.\u00a0\u2013\u00a0@jessesem<\/span><\/a><\/p>\n

we define it as finding who\/what contributed to a conversion\/traffic\u00a0\u2013\u00a0@JonKagan<\/span><\/a><\/p>\n

In regards to Paid ads, “attribution” is whichever way the conversion was attained (i.e., if it was a keyword higher in the funnel or lower in the funnel).\u00a0\u2013\u00a0@marccxmedia<\/span><\/a><\/p>\n

 <\/p>\n

 <\/p>\n

Q2 Is attribution a metric you currently use to determine the success of different paid ad channels? Is so, how do you use it?<\/strong><\/h4>\n

 <\/p>\n

Atribution is a best guess at the source of your conversions. It’s not foolproof, nor is it gospel, but done right it can make you a much smarter marketer\u00a0\u2013\u00a0@armondhammer<\/span><\/a><\/p>\n

At both agencies I was at, we used last-click as a “yes, this isn’t great, but it’s consistent” method. Working with businesses, though, it was challenging to align our attribution to their internal counts.\u00a0From what I can grasp — the *right* way to do attribution is by modeling based on revenue and spend. But most agencies don’t have the data to do that, and most businesses have better things to spend time on.\u00a0 \u2013\u00a0@ferkungamaboobo<\/span><\/a><\/p>\n

Yes when reporting on ROAS I like to show the difference in ROAS between different attribution models.\u00a0\u2013\u00a0@StephanieErne<\/span><\/a><\/p>\n

Doing marketing without attribution is driving without a seatbelt. Sure you can do it, but it’s not smart. We always try to have some sort of success metric for any spend and tie back to it. \u2013\u00a0@armondhammer<\/span><\/a><\/p>\n

If you aren\u2019t using attribution I\u2019m not sure how you can have success (unless you\u2019re just getting VC and don\u2019t care about CAC). I think the focus on attribution has really shed light on a need for multi-channel strategies and creating a customer journey\u00a0\u2013\u00a0@markpgus<\/span><\/a><\/p>\n

Yep. But it is for both paid and non-paid.\u00a0\u2013\u00a0@JonKagan<\/span><\/a><\/p>\n

We tend to determine success in Paid channels based on how many conversions (form fill-outs and calls) we’re getting, what keywords are performing best, etc.\u00a0\u2013\u00a0@marccxmedia<\/span><\/a><\/p>\n

 <\/p>\n

 <\/p>\n

Q3 Which attribution model in Google ads do you use the most and why? Last Click, First Click, Linear, Time Decay, Position based or Data-Driven?<\/strong><\/h4>\n

 <\/p>\n

We tend to believe the models of GA more than adwords, and prefer linear models. It still leaves out impressions without clicks and the like so not perfect, but at least it considers intro to the site as valuable. \u2013\u00a0@armondhammer<\/span><\/a><\/p>\n

We tend to stick with last click since you can’t change the default in Google analytics.\u00a0\u2013\u00a0@elevatedmrktng<\/span><\/a><\/p>\n

In Google Ads it depends on the client and purpose of his \/ her campaigns. But I often use linear. Would be great to be able to set this per campaign instead of on the account level\u00a0\u2013\u00a0@StephanieErne<\/span><\/a><\/p>\n

Data-driven. Because, YouTube.\u00a0\u00a0\u2013\u00a0@jessesem<\/span><\/a><\/p>\n

IMO this depends 100% on the client. Look at Time Lag report and path length. However, Data-Driven and position based are used most. Time decay can be powerful in certain circumstances.\u00a0\u2013\u00a0@markpgus<\/span><\/a><\/p>\n

Linear and last click. I want to know what did the converted, and what else helped.\u00a0\u2013\u00a0@JonKagan<\/span><\/a><\/p>\n

I believe we tend to use Last Click as the default.\u2013\u00a0@marccxmedia<\/span><\/a><\/p>\n

 <\/p>\n

 <\/p>\n

Q4 Which attribution window combination do you use the most in Facebook ads and why? The default (1-day view and 28-day click) or do you change it? If so what do you use and why?<\/strong><\/h4>\n

 <\/p>\n

I just use the default. Just because we don’t run very large campaigns that generate loads of revenue in the end.\u00a0\u2013\u00a0@StephanieErne<\/span><\/a><\/p>\n

I tend to stick with the default for reporting and will change it only when diagnosing issues and analyzing performance.\u00a0\u2013\u00a0\u00a0@jdprater<\/span><\/a><\/p>\n

I feel like with attribution my answer is always going to be \u201cIt depends\u201d If the objective of a campaign is an LP view and retarget elsewhere, that\u2019s going to be different than if you\u2019re doing bottom funnel retargeting. It\u2019s important that the window has a purpose\u00a0\u2013\u00a0@markpgus<\/span><\/a><\/p>\n

I use the default, in part because everything tracks differently. I’m a big fan of where your different tools say different numbers — it tells you more than if everything says the same thing.\u00a0\u2013\u00a0@ferkungamaboobo<\/span><\/a><\/p>\n

It has been a little while since we ran a Facebook Ads campaign, so I don’t remember what our last attribution window combination was. I think we set it pretty wide, though.\u2013\u00a0@marccxmedia<\/span><\/a><\/p>\n

 <\/p>\n

 <\/p>\n

Q5 What challenges do you have when it comes to multi-channel (for example Facebook vs Google ads) attribution and how have you overcome those?<\/strong><\/h4>\n

 <\/p>\n

FB (and most display) tends to be greedy with the attribution. We’ve had to flight ads on and off to get a feel for real lift.\u00a0\u2013\u00a0@armondhammer<\/span><\/a><\/p>\n

I tend to use GA for measuring website performance and FB Analytics for measuring user performance.\u00a0\u2013\u00a0\u00a0@jdprater<\/span><\/a><\/p>\n

We struggle with clients understanding full funnels, that all marketing affects the other.\u00a0\u2013\u00a0@elevatedmrktng<\/span><\/a><\/p>\n

They just never add up. Best is to know exactly how things are measured \/ determined so you can explain it clearly. I tend to use simple examples preferably tailored to someone I am talking to. But the basic thing is fully understanding it yourself I think.\u00a0\u2013\u00a0@StephanieErne<\/span><\/a><\/p>\n

When Youtube is introduced\u2026 kinda difficult. Impossible to attribute those 29 sec viewers too. Same with anything view-through actually. Really difficult.\u00a0\u2013\u00a0@markpgus<\/span><\/a><\/p>\n

Sometimes the multi channel tools hit an indecipherable firewall that slows us all down\u00a0\u2013\u00a0@JonKagan<\/span><\/a><\/p>\n

When it comes to attribution, we tend to stick to single ad platforms (AdWords campaign only, Facebook campaign only) but in the event of some potential upcoming campaigns across platforms, UTMs will help differentiate traffic (which ties to Q8).\u00a0\u2013\u00a0@marccxmedia<\/span><\/a><\/p>\n

 <\/p>\n

 <\/p>\n

Q6 Since attribution isn\u2019t perfect, how do you weigh its importance to other metrics you look at?<\/strong><\/h4>\n

 <\/p>\n

mixed bag, i hate to do even weight, but it is often something i am forced into\u00a0\u2013\u00a0@JonKagan<\/span><\/a><\/p>\n

The worst attribution I’ve had was a multi-device long cycle. Read info on phone, ask boss, convert (later) on desktop. 90% direct. Had to use engagement for attribution. Did they look at pricing after another key page? It’s still some form of attribution tho\u00a0\u2013\u00a0@armondhammer<\/span><\/a><\/p>\n

The same I think. So many data is corrupted. Fake likes, you’ve got adblockers, cookiebars that block stuff, new spam popping up. You never get it 100% right. In the end profit (or cost \/ revenue) matters most. When you go granular you just use data the best you can.\u00a0\u00a0\u2013\u00a0@StephanieErne<\/span><\/a><\/p>\n

We tend to focus more on the conversion-related metrics rather than attribution. \u2013\u00a0@marccxmedia<\/span><\/a><\/p>\n

 <\/p>\n

 <\/p>\n

Q7 How does attribution affect your budgeting across multiple paid channels? How often do you “shift” money from one channel to the other based on attribution<\/strong><\/h4>\n

 <\/p>\n

it doesn’t really affect budgeting for me. I tend to look at results, goals, and scalability when shifting money around\u00a0\u00a0\u2013\u00a0\u00a0@jdprater<\/span><\/a><\/p>\n

helps us determine if we need to reallocated a higher amount at the top of the funnel or not. \u2013\u00a0@JonKagan<\/span><\/a><\/p>\n

I look more at funnel stage than channel. I assign channels to stages of the funnel. Throw as much budget lower funnel as makes sense and slowly go higher making sure not to be spread too thin. Always remember to optimize the traffic coming in for free first!\u00a0\u00a0\u2013\u00a0@markpgus<\/span><\/a><\/p>\n

Scale. So key. You can get all of the 3000% roas you want on brand search, but how do you get more? You spend off line or on that “lagging” display\u00a0\u2013\u00a0@armondhammer<\/span><\/a><\/p>\n

Sometimes (but not often) I do make choices on this when there is a big difference in performance and a small budget. First test a some things & then decide where to best spend the money. Or as I say where the chance is largest that you get the most value out of it.\u00a0\u2013\u00a0@StephanieErne<\/span><\/a><\/p>\n

As mentioned in A6, we tend to focus more on conversions rather than attribution, so we shift budget around based on how campaigns are performing conversion-wise.\u00a0\u2013\u00a0@marccxmedia<\/span><\/a><\/p>\n

 <\/p>\n

 <\/p>\n

Q8 What tools do you use to help with attribution?<\/strong><\/h4>\n

 <\/p>\n

due to budgetary restrictions, we rely heavily (way to heavily) on free ones like GA\u00a0\u2013\u00a0@JonKagan<\/span><\/a><\/p>\n

We do not use any specific tools besides the advertising platforms and Google Analytics. We do want to use call-tracking more. That also counts when talking attribution right\u00a0\u00a0\u2013\u00a0@StephanieErne<\/span><\/a><\/p>\n

 <\/p>\n

 <\/p>\n

Q9 How are you currently using the Google ads attribution tool?<\/strong><\/h4>\n

 <\/p>\n

-NA-<\/p>\n

 <\/p>\n

Q10 What is your main platform for showing data for attribution? (for example, do you use Google Analytics or another platform?)<\/strong><\/h4>\n

 <\/p>\n

Custom reports using Looker or Tableau. Use GA if you don’t have a team to pull off the APIs.\u00a0\u2013\u00a0@jessesem<\/span><\/a><\/p>\n

GA because I am a GALover\u00a0\u2013\u00a0@JonKagan<\/span><\/a><\/p>\n

 <\/p>\n

 <\/p>\n

Q11 Do you use and talk about attribution by channel with your clients?<\/strong><\/h4>\n

 <\/p>\n

yes, with every client\u00a0\u2013\u00a0@JonKagan<\/span><\/a><\/p>\n

 <\/p>\n

PPCChat Participants:<\/strong><\/p>\n