PPC Chat Streamcap – Taking a Closer Look at Cost per Conversion

This week Matt Umbro (@Matt_Umbro) came up with this question set titled “Taking a Closer Look at Cost per Conversion.” The following is the transcribed Streamcap from the live chat:

Q1: How do you & your clients determine target cost per conversions (if you do at all)?

  • We have the client set a ceiling where they STOP making $. Avg. Sale * lead-to-sale rate = CPA goal. If we’re under, we win! – Luke Alley (@LukeAlley)
  • Ideally CPA is a % of lifetime customer value. not always possible though. – Harris Neifield (@Harris Neifield)
  • Depends on ecommerce/sales or lead gen. Lifetime value works it’s way into ecom. depending on expected repeat vs. 1 time sales. – James Svoboda (@Realicity)
  • Non-branded term target CPC’s should be based on customer lifetime value. Branded term CPC’s can be based on transaction ROI. – Tracy Henry (@tracy_a_henry)
  • Our goals are all ROI based, with sitelinks rolled into the performance of the segment they appear in.- Chris Kostecki (@chriskos)
  • Look at profit margins and cost across product sets and set target CPA based on revenue and ad spend goals for each product set. – Jessica Cameron Ruud (@Camruud)
  • I think the key here is Customization. Every brand/site is different. – Emily Las (@emlas)
  • Depends on the type of site, but for eCommerce: generally 10% of avg. basket size. – Ravi Sodha (@ravisodha)
  • Ideally, % of customer lifetime value for generic terms, % of transaction value for branded terms. – Tracy Henry
  • (1) Arbitrarily/POMA; (2) based on baseline data; (3) back into it using customer LTV calculations. – Melissa Mackey(@Mel66)
  • First step is answering the question: What is this conversion worth to the business? Then define "allowable" CPA. Sometimes I define separate max cost per conversion for new vs. returning customers. – Emily Las

Q2: How has ad positioning affected your cost per conversions?

  • I generally aim for position 3 to start, then adjust bids (and position) based on relation to the CPA goal. – Aaron Levy (@bigalittlea)
  • Strangely, pos 1 has lower CPAs than 2-4. Maybe better trust? No idea. But I don’t hate it. – Michelle Morgan (@michellemsem)
  • Higher converting terms can afford to be in higher position. More Conversions = lower cost per conversion. – Jessica Cameron Ruud
  • Top pos. ads can be there due to high conv. rates, so they can have low CPA’s. Surprises some (higher pos. – pay more) but true. Chicken & egg prob. When the conv. rate is high, CPA is low & bids are upped to raise volume. High pos. looks good for CPA. – Harris Neifield
  • It takes a pretty massive data set to prove statistical significance. I either haven’t gotten there or there is no difference. – Tracy Henry
  • Ad data matched to irrelevant KWs that eventually become negatives have larger impact that AVG Pos. – James Svoboda
  • "Other" ads have 5-10x higher cost/conv than Top ads. Aside from that, there’s little variation. – Melissa Mackey
  • Depends on the product, level of involvement in purchase, etc. sometimes pos 1 is best, others 3, etc. – Alma Smith (@Alma_Smith)
  • Depends on vertical. In comp. niches seems that other positions do better than top. Maybe people fill forms for multiple advert. – Luke Alley 
  • Conv. rates by position are a chicken & egg question (PPCers bid up when conv. rate is high) unless you run an ACE position test. – Harris Neifield
  • Bottom line – If you have strong copy and landing page cost per conversion will naturally decrease no matter the position the caveat being that your ads are getting enough exposure. – Matt Umbro (@Matt_Umbro)
    •  Exactly. I don’t think it’s coincidence that higher spots convert better. More $$ for Goog this way. In other words, all the bling for the top spots doesn’t just help CTR… Helps conv. rate too. – Luke Alley
      • I’ve found cost per conversion is similar between top & side. It’s CTR that differs greatly. – Matt Umbro
  • That’s a hard question because it differs w/industry. Some work best in 1st – 2nd pos and some in 5th – 6th. You have to test. – PurePPC.com (@pureppccom)
  • I find it kinda dangerous to bid to target position. If keyword does well, bid up. Does poorly, bid down or change! – Aaron Levy
    • Agree, we can’t lose sight of CPA by focusing too much on position. bid up or down if applicable unless running a test. – Harris Neifield
  • Hard to tell because you would have to do a SERP look up and see if you have true competition above and below. – James Svoboda
  • In some instances I have seen CVR (ultimately affecting CPA) increase as ad position has decreased – but it depends a lot on the KW. – Atalley (@atalley)
  • Great point– Sitelinks can provide an extra boost in top spot. Closer to point-of-purchase. – Emily Las
  • Top spot definitely has the most variations to offer as far as ad format and extensions. – Michelle Morgan
  • It also depends on how you are valuing different types of convs, whether they are 1-per-click or many, mobile v.desktop etc. – Atalley

Q3: Is cost per conversion the primary metric you optimize for? Why or why not?

  • Volume is also a big factor. If cost-per-conversion goes up, but is reasonable, and volume goes up, it’s a win win. – James Svoboda
  • Cost per Conversion is not our "end all be all" metric since we utilize call tracking within our lead gen. strategy. – Nicole Mintiens (@Tregesy)
    • Good point on call tracking. Sometimes that, and assisted conversions, don’t clearly reflect in conversion data. – James Svoboda 
  • Yes – our clients invest in PPC for a business benefit (the conversion). – Harris Neifield
  • Depends on the vertical/conversion goal I’m working with. Cost per is not typically the focus. More about volume, ROI. – Emily Las
  • No, cost per conversion is a top metric but ROAS is most important metric. – Matt Umbro
  • It’s right up there with # of conversions. Have to weigh how many conversions with CPA. They play off of each other. – Michelle Morgan
  • ROAS is what we focus on as top priority. – Gnosis Media Group (@GnosisArts)
  • Cost per conv is key for B2B but as James Svoboda says, sometimes volume is priority. Best to find a blend of both. – Aaron Levy
  • I think the combination of all of the metrics, if you know your baselines, is the best measurement of acct/campaign performance. – Atalley
  • We primarily work with ecomm clients where ROAS is most important. We also track revenue down to the keyword level. – Matt Umbro
  • I think the combination of all of the metrics, if you know your baselines, is the best measurement of acct/campaign performance. – Atalley
  • As others have been saying, cost per conv is kinda irrelevant for e-comm IMO. Rather focus on ROAS & pay more for big orders! – Aaron Levy
  • ROAS is good when a client has multiple products and or different margins by product (often e-comm). – Harris Neifield
    • Good point, cost per conversion is still huge factor but if return is there CPC can be higher. – Matt Umbro
  • Using ROAS as the top metric. – Alma Smith

Q4: In what situations is a high cost per conversion OK? 

  • When the value of the return is higher. – Francis Shovlin (@fmshovlin)
  • When overall conv volume is low; when product cost is large; when you’re focused on brand/imp share. – Melissa Mackey
  • Personal injury lawyer in Los Angeles. – Luke Alley
  • If you’re working in Insurance. – Paul Kragthorpe (@PaulKragthorpe)
  • I’ll pay a high cost per conversion if we’re promoting a big money, high margin products so it’s worth it! – Harris Neifield
  • When it’s a higher quality conversion (Search > Content, Keyword A > Keyword B). – Michelle Morgan
  • When a lead (or sale) is worth more 🙂 If kwd A leads to average sales of $1k and kwd b leads to sales of $10k, I’m paying for B. – Aaron Levy
  • When customer becomes a repeat customer. Considering the customer lifetime value. – Chris Smeda (@ChrisAlanSmeda)
  • If cost-per-conversion is high and still reasonable, and volume it acceptable. Also factor in 2nd tier convs. like Email subs. – James Svoboda
  • 1 of our client’s is a hotel in nyc @ $500+ / night. EZ to justify high CPC. Essentially, when the product/service has high yield profit margin. – Gnosis Media Group
  • When the focus is on brand awareness and/or for highly competitive terms. – Atalley
  • IT related keywords are huge and tend to have a long sales cycle, but when they convert they pay for that year’s PPC budget. – Matt Umbro
  • A high cost/conv is ok if proper attribution is considered. example: top of the funnel terms/ad groups = fewer attributed conv. – James Zolman (@jameszol)

Q5: Do you report many-per-click cost per conversions? If so, what insight or explanation do you share with clients?

  • Nope, not a many-per-conversions type of girl. (Whatever that means). – Michelle Morgan
  • If multiple conversions have value, we report on MPC (e-comm, multiple actions). Otherwise, it’s OPC. – Aaron Levy
  • I have only reported many-per-click for eComm clients. Want to only include unique leads for B2B. – Francis Shovlin
  • Only report when relevant (ie when multiple purchases might have been made in last 30 days). – Robery Brady (@robert_brady)
  • Yes, w/total value per conversion & explanation of multi-touch conversions. – Melissa Mackey
  • Yes, when multiple-step forms are involved. Explain the difference in each. – Luke Alley
  • I report many-per-click conversions and metrics with caution, sometimes are a result of visitor hitting refresh on checkout page. – Matt Umbro
  • This is where I like using 3rd party software. I don’t like how Google records conversions to click date rather than conv date. – Emily Las
    • Especially when Google Analytics records conversions on conversion date and NOT click date. – Matt Umbro
  • No. Many often pulls in less important conversions like email subscribers when sales conversions are the key data point. – James Svoboda
  • Many per click is useful for e-comm but if you’re counting leads or using a customer LTV model, 1-per-click is better. – Harris Neifield
  • I like to test by checking both 1-per-click & many-per-click against the client’s data from back end. Then go from there. – Emily Las
  • I only tag things that I 100% want to report as convs. Track other stuff in analytics as "nice to know" metrics. – Aaron Levy
    • Sometimes need to track engagement metrics in AdWords to help determine better keywords, ads, & ad groups. – James Svoboda
      • Agreed, I’ll usually set the end goal as the conversion with a higher CPA goal to account for it. – Aaron Levy
  • depends. For ecommerce websites I do, since customers can buy multiple items in multiple sessions. – Bart Schuijt (@BartSchuijt)

Q6: How do you predict the implementation of extension level conversion metrics will influence your management and reporting?

  • One more way to demonstrate how awesome you are. Shows additional granularity. – Robert Brady
  • It probably won’t. We expect reports to travel and want to keep them high level so the C-Level can understand them management will become different, but not reporting. – Harris Neifield
  • Not much until they start matching extension data to Ad Groups, Ads and Keywords. – James Svoboda
  • You could always see extension conversion metrics by segmenting by click type, but not it is more convenient. – Matt Umbro
  • Reports will stay the same; most don’t need to know that much detail. But, may be good in back end in case client wants to dig. – Aaron Levy
  • Really need conversions by extension & also ability to pause & test to really make reporting on them worthwhile. – Melissa Mackey

Q7: Do you break out cost per conversions by conversion type or report at aggregate level? Why?

  • Yes. We do both. A rollup for top-level metrics and breakdown to show trends by conv type. – Melissa Mackey
  • Aggregate. Impossible to "attribute" spend to conversion A vs. Conversion B unless campaigns are silo’d. – Aaron Levy
  • Both, but mostly by conversion type as they have different values. – James Svoboda
  • I always break out type. Especially because at least 1 type will have revenue associated- Can’t skew that AOV! – Emily Las
  • Wouldn’t make sense to break out (would inflate CPA for each lead type). break out to see what is driving the most. – Francis Shovlin
  • Also depends on the type of conversion. An email signup isn’t worth as much as a sale. – Melissa Mackey


More PPCChats

Don’t forget to stay tuned for the next #PPCchat on Tuesday at 12 noon Eastern, 9 am Pacific and 5pm in the UK. Same Chat time, same Chat channel.

CRO Chat

Announcing CRO Chat: A Weekly Open Discussion About Digital Conversions. Thursdays at 12 noon Eastern, 9 am Pacific.


Check out the PPCChat Twitter list to see and connect with all current and prior participants.

About the Tiger Woods of Streamcaps

This is a guest post by Paul Kragthorpe; PPC Search Manager for WebRanking in Minneapolis, Minnesota, #PPCChat Streamcap Grabber, Blogger. Connect with me @PaulKragthorpe, and Google Plus.

Be Sociable, Share!

Tags: , ,

Leave a Reply

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

Recent Tweets

I am speaking at SMX East
PPCChat.co was rated one of the Best PPC Blogs by Boost CTR