Paid Search and ROI

March 28, 2008 - By Laura Callow

Paid Search is not about ROI. Technically, it’s not about Return on Ad Spend (ROAS) or conversions either. Paid Search should be about profit. Excuse the SEMI version of Finance 101 here ;) but it’s for a reason. Bear with me…

• ROI = Profit / costs x 100
• Profit = Revenue – Costs

Companies are interested in ROI as it is a percentage return based on capital invested. But Investment in alternative advertising media can by no means provide you with a realistic ROI as the ad spend utilised in the ‘cost’ field does not take into account the product development, manufacturing, distribution and other associated expenses.

Another point to take into consideration is that ‘ROI’ for paid search text ads is often referred to as Return on Ad Spend (ROAS). It would far better be referred to as Return on Ad Investment (ROAI) as opposed to ROAS which can be easily misconstrued to refer to a simple CTR x bid-value amount.

In the world of finance, the lower the ROI the less risky the investment, and the higher the ROI the more risky the investment. This is not true for paid search which depends on bid-for-placements.

The more you bid, the less likely it is that you are risking not getting CTR or conversions. The greater your CTR (in general with a relevant, usable web site) the greater your conversion rate is likely to be. The degree to which revenues increase with cost expenditure is not nearly as linear as one would expect from other marketing media investments due to the auction type nature of these bids, and the immense discrepenacies both in click-through rates and conversion rates across industries simply between position 1 and position 10 in the paid search results.

The following table provides an example that will explain this in a more visual manner. It is hypothetical, but illustrates the point.

roi-paid-search-profit.jpg

The most sensible AdSpend here is not the one which maxes on CTR, nor the one which maxes on gross revenue. Neither is it the option which maxes on ROAI or conversions. It is the one which provides the maximum profit for this particular advertising medium.

This profit can then be applied to the actual costs of product development, production, distribution and associated costs to allow for a true measurement of ROI.

Make your paid search work for you by analysing it in terms of campaign profit. If you do, the results and eventual impact on ROI will be tangible and well worth the ad spend.

Spread the word:

We'd be honored if you'd help support SEMI by Stumbling us or voting for us on Sphinn.


Stumble it!

5 Comments so far... perhaps you would like to leave one?

Laura great post! I really like how you broke it down in 101 style terms. I am going to use this post to try and keep my clients in check when they start asking about AdSense and ROI! Thanks again. And I really look forward to more of your posts! Sphunn!

http://sphinn.com/story/37543

Comment by Martin Bowling — March 28, 2008 @ 7:37 pm

I had easy time reading your blog. But it seems now it’s over :( . Man, this post sucks. I hope at least the next one won’t be.

Comment by Matt — April 10, 2008 @ 9:42 pm

Sorry you didn’t like it Matt. Lucky for me it got quite a bit of positive feedback on Sphinn :)

Comment by Laura Callow — April 11, 2008 @ 4:48 am

Laura, I like the fact that you have included management fees in your analysis, a point too often ignored by Ad Managers and Advertisers. However, I would have characterized the setup fees as a true investment, and set that aside entirely, or at least amortize it over 12 months.

Aside from that, I disagree with you on several other points:

Presumably your “CTR” column represents clicks, not CTR.

Your hypothetical case doesn’t apply to some accounts. For example, our experience shows that any online store catering to price sensitive shoppers almost always performs best below the fold, in positions 5 to 8.

Ad Managers frequently create hypothetical cases which ignores reality. The reality is that most Advertisers don’t buy 100% of available impressions.

Your analysis can only be applied if one understands the account in the context of impression share and budget constraints, and what alternative opportunities exist for each element of the account. For example, if the above account is achieving a 20% impression share, and losing 50% of impressions due to budget, then a lower bid might actually result in more traffic, not less traffic, and since we know that the below the fold positions favor shoppers, if Advertiser’s prices are very competitive, then below the fold bids might actually yield MORE traffic and revenue, not less traffic and revenue (because the budget / lower CPC is higher when the CPC is lower).

Comment by Chris Lude — November 3, 2008 @ 6:13 am

Thanks Chris,
“However, I would have characterized the setup fees as a true investment, and set that aside entirely, or at least amortize it over 12 months.”
…For sure – these are fixed (or sunk) costs, and as with any investment, they do need to be taken into consideration, if simply to assit in initial budgeting. This was an illustrative post, not an ecos or accts 101 post.

“Aside from that, I disagree with you on several other points:”
…Your prerogative :) – always appreciate readers getting involved.

“Presumably your “CTR” column represents clicks, not CTR.”
…You are absolutely correct. My apologies. Please consider for this hypothertical case that as bid value increases, click throughs increase, but not linearly.

“Your hypothetical case doesn’t apply to some accounts. For example, our experience shows that any online store catering to price sensitive shoppers almost always performs best below the fold, in positions 5 to 8.”
…Thank you for that insight – and yes I realise that this hypothetical blog post cannot even begin to apply to every possible scenario. It does however seem to do a relatively good job of illustrating the primary point that profit needs to be taken into consideration and not simply cost per conversion.

“Ad Managers frequently create hypothetical cases which ignores reality. The reality is that most Advertisers don’t buy 100% of available impressions.”
…I don’t believe this hypothetical case ignores reality. Being an hypothesis, it is conjectural and is simply an attempt to provide some guidance or insights based on sumise rather than hard evidence. Again, the point I was trying to impart was that (from a 101 perspective) PPC advertisers should not get hooked on minimising their cost per conversions without taking into account actual profitability. The number of impressions advertisers buy is dependant on a number of variables. That too is beyond the remit of this post.

“Your analysis can only be applied if one understands the account in the context of impression share and budget constraints, and what alternative opportunities exist for each element of the account. For example, if the above account is achieving a 20% impression share, and losing 50% of impressions due to budget, then a lower bid might actually result in more traffic, not less traffic, and since we know that the below the fold positions favor shoppers, if Advertiser’s prices are very competitive, then below the fold bids might actually yield MORE traffic and revenue, not less traffic and revenue (because the budget / lower CPC is higher when the CPC is lower).”
…I am not disputing this point, but the detail you would have liked me to go into from what I can tell from your comment is too advanced for the nature of this post, and while it is intersting, and worthy of a post of its own, it neither adds to nor detracts from the point of THIS post – bear in mind campaign profitability, and don’t get stuck on CPC data when determining your ROI…

Thanks!

Comment by Laura Callow — November 3, 2008 @ 9:46 am

RSS feed for comments on this post. TrackBack URL

Leave a comment