What’s A Conversion?
A conversion is a desired action or series of actions that a prospective consumer can make, indicating a progression through the buying cycle. The most obvious type of conversion is the actual exchange of money for goods & services. When a consumer enters a store and buys something, the store has ‘converted’ that prospect to a customer. In the ecommerce world, when a consumer progresses through the online checkout process by giving their credit card information and shipping address, the ecommerce store has also made a conversion.
For many businesses, tracing how a customer
- 1) first got to their website,
- 2) navigated around that website,
- 3) then eventually visited the store
is difficult to do. How can one optimize a site or optimize an online ad campaign if there is no clear indicator of what tactics work?
In this case, understanding how a customer moves through the buying process using free tools like Google Analytics for a given niche & then designating a specific “indicator” on a website as a “conversion” provides marketers better statistics on what works and what doesn’t.
In the case of a local retailer, defining the desired conversion could mean something as simple, but telling, as: “a website visitor clicking on the ‘Contact Us’ page”. A prospect’s move toward a website page that provides a business’ phone number, contact e-mail, hours of operation, & physical store address is an indicator that the prospect is moving further in the buying cycle towards becoming a customer.
Cost Per Conversion
What is ‘cost per conversion’? Simply put:
The amount of money spent to create a conversion over the number of conversions actually made.
Let’s use a local chiropractor as an example.
This fictitious chiropractor knows, through meticulous record keeping, that he provides services for an average of 20 clients per month that found his chiropractic office through the internet. He also knows, by checking with is his favorite internet marketer, that 40 unique website visitors view his ‘Contact’ page per month. Additionally, he knows that his site generates 400 unique visitors per month directly attributable to his fictitious $500/month SEM budget.
What does this tell him?
1) 50% of online conversionsIn this case, visiting the website's 'Contact' page. ultimately visit chiropractor
2) 10% of unique website visitor make an online conversion
3) Cost of “Actual” Conversion for chiropractor is ($500)/(20 clients) = $25
What then do we ask?
“Would I, as this fictitious chiropractor, pay $25 to guarantee a new client?” It depends.
It depends on the revenue generated by this new client.
It depends on the profit margin from that revenue.
It depends on my willingness to provide services at normal margin minus cost of conversion.
Conversions are often mistakenly measured as a percentage of traffic to a desired action (ie. 8% of website ‘visitors A-Z’ completed ‘action #1’), and often fail to measure the cost of attracting ‘visitor A’ vs. ‘visitor B’ vs. ‘visitor C’ vs. … ‘visitor Z’. If A & B convert at the same rate, but A only cost $.50 to get to the website & B cost $1, it stands to reason that money spent advertising to B should be reallocated to A.
A/B Split & Multivariate Testing
Fortunately, there is more to ask than just: “Does this model provide a positive and sufficient ROI?”
What conversion tracking allows you to ask is: “How can I increase ROI by decreasing cost per conversion?”
With today’s analytics tools, it’s very possible to track:
- How many web surfers are visiting a site
- How long users are staying on the site
- Which search engine a visitor used to find the site
- Whether these visits were paid for vs. organic
- What search keywords a visitor found the site by
- What landing pages those search queries brought the visitor to
- What the bounce rate (the percentage of visitors who leave a website at the first landing page, and do not continue browsing through more of the site) for those landing pages are
After a while, the incentive to measure these statistics becomes overwhelmingly obvious. If one can track the variables that lead to more conversions, then manipulating a website or ad campaign to maximize these conversions becomes fairly straightforward.
Which keywords lead to more conversions?
This is your first, most infant, step to optimizing a successful PPC campaign. Let’s say our fictitious chiropractor is running an SEM campaign for three keywords:
- “Chiropractor in [insert city]” – Average CPC $2.25
- “Back Pain” – Average CPC $.65
- “Spinal Pain” – Average CPC $.65
Our chiropractor knows that, like in the previous example, 50% of website visitors who make an ‘online conversion’ ultimate convert into clients. He also knows that he is not willing spend more than $55 per new client (ie. cost per ‘actual conversion’).
According to analytics, the online conversion rate for:
- “Chiropractor in [insert city]” is 20%
- “Back Pain” is 2%
- “Spinal Pain” is 2.5%
Keywords: Chiropractor in [insert city]
[ ($2.25 CPC) / (20% conversion rate) ] / (50% online:actual conversion ratio) =
Keywords: Back Pain
[ ($0.65 CPC) / (2% conversion rate) ] / (50% online:actual conversion ratio) =
Keywords: Spinal Pain
[ ($0.65 CPC) / (2.5% conversion rate) ] / (50% online:actual conversion ratio) =
Although the keyword phrase “back pain” is less CPC than is “chiropractor in [insert city]”, the cost per conversion is still lower. Although the CPC of “back pain” and “spinal pain” are identical, the chiropractor is not willing to allocate advertising budget to only the former due to cost per conversion.
Which geographical areas display a higher conversion rate?
The key here is to set up multiple advertising campaigns with:
- Identical Ad GroupsA subcategory of a campaign often grouping similar keywords. All keywords triggered by ad group ‘Z’ will only trigger text ads related to ad group ‘Z’.
- Identical Keywords
- Identical BidsThe maximum an advertiser is willing to pay for a web surfer to visit the website for a particular search query for those Keywords
- Identical Text AdsThe visible text associate with a listing in the SERPS—comprised of link anchor text, the destination URL, and a brief description of the website
- Identical Landing Pages
- Different demographic & geographic locations
Displaying a chiropractic ad in a metropolitan area may produce more leads, but do those leads come at a higher cost per conversion? Increased population could mean increased competition. Increased competition means more chiropractic businesses bidding for the same scarce advertising space.
After a statistically significant sample size is collected, maybe our fictitious chiropractor realizes that his ads over the large, nearby retirement community are producing the lowest costs per conversion. Maybe, conversely, the costs per conversion around the University’s student housing are too high.
There is significant waste if a marketer isn’t taking the time to run these sorts of split tests, and is instead running a simple, blanket campaign that covers all of a business’s service area.
A successful Search Engine Marketer allocates an advertising budget based on records and statistics, not just on intuition and gut.
How can varying content on my site affect conversion rates?
Not only can we split test differing keyword phrases and text ads for conversion, but A/B testing different landing pages is also a great (although sometimes cumbersome) first step to optimizing the user experience from an ‘on-site’ standpoint.
Catching ‘Red Flags’
After enough data is collected, it’s easy to find the campaign variations that aren’t working. Did you find a text ad with a below average CTR? Cut it and split the remaining ad(s) into similar, but unique, variations. Have a keyword phrase that just isn’t performing? Perhaps a look at your competition will tell you that the bid on a chosen keyword is too steep to turn a profit, or that competitive saturation in the market for goods/services related to that keyword phrase is weeding out profit margin.
Suspicious Bounce Rates
Content or design layout errors on landing pages may also turn prospective clients off to your product. Landing pages with sub-par design layout, broken images, broken links, or poor content will ultimately display higher bounce rates. Routinely checking for pages with high bounce rates will ensure that coding errors within the site are reduced and that landing pages with poor content are reevaluated.