This is part of a series on Wine and Landing Page Optimization (LPO). The previous post outlined why you should do LPO. This post will look at the first steps of how to go about testing. A later post will look at installation and the LPO process.
Step 1 Set up Google Analytics
If you don’t have Google Analytics installed then it’s probably better to do this first and then come back to Website Optimizer – “walk before you run” kinda thing.
It is also best if you have the eCommerce option selected. Depending on your website software, installing basic Google Analytics is usually easy. The eCommerce option can be tricky though and will probably require some expertise. The eCommerce option is important for tracking profit rather than just traffic.
Step 2 Set a Baseline
You need to set a baseline, or a benchmark, of your current results. The aim will be to attempt to beat this benchmark time and again with the better test becoming the new benchmark, thereby ratcheting the performance of your page up. You may also find your new tests don’t do as well as the benchmark so you end up discarding them and stay with the benchmark.
In the direct marketing industry this is called “beating the control”.
In order of preference:
- Profit per visitor
- Revenue per visitor
- Conversion rate
- Shopping cart abandonment rate
========= Sidebar on Calculations =========
Conversion rate = visitors who purchase pa / total number of visitors pa
Revenue per visitor = average revenue per sale / total number of visitors
Average revenue per sale = (price * quantity * % first time purchases) + (price * quantity * % repeat purchases)
Note repeat customers tend to buy more at higher prices.
Average profit per sale = (price – COGS) * quantity – credit card fee – average net freight cost
In order to take out seasonality influences you usually use time periods such as per annum (pa), or versus the same period last year.
You should use Profit per visitor.
This requires Google to have data on the margins of your wines which may be technically challenging. It’s is usually easier to provide access to prices. If that’s the case then Revenue per visitor will do. The problem is you want to increase profit not revenue – you could increase revenue but reduce profit by cannibalizing high margin wines.
If you’re unable to access prices then fall back on conversion rates and shopping cart abandonment. Benchmark your these as follows (the % are best practice from Practical Ecommerce):
- Visit to Product page Ratio (60%)
- Product page to Cart Ratio (30%)
- Cart to Checkout Ratio (60%)
- Checkout to Order Rato (75%)
Basic measurements of success all set up? Let’s move on to testing.
Step 3 Decide what to test
Just choose two simple things to get started and then get more sophisticated over time.
Here’s what one expert firm did with their own landing page:
The usual suspects:
- Product description copy
- Offers (e.g. free shipping, 1 case only)
- The “call to action” (e.g. Buy Now vs Add to Cart, button at top or bottom or middle, high contrast, different colours, “Register” vs “Delivery Details”)
- Clean design – get rid of as many links and other distractions as possible, focus on the the key action for the page usually the product purchase
- Ratings and medals (have them or not)
- Reviews (have them or not)
- Prices (test price points)
- Single Page Checkout vs multi page (hint: go single)
- Guest checkout option vs forced registration (hint: Guest is Best)
Gather testing feedback, conjecture, insults and opinions by:
- 5 or 10 sec test – informal (show your mother/acquaintance your website for 5 seconds and ask them what the page is about)
- Usability tests – paid see CRO – Five Second Test, Crazy Egg, UserTesting.com to which I would add clicktale.com
- Online and Email Surveys
- Feedback requests on website, like Feedback boxes you see on the side of pages (like mine!)
- Team brainstorming
- Look at your competitors sites
If you use surveys here are some questions you might ask:
- “How likely is it that you would recommend us to a friend or colleague?”
- “What other wine regions, varietals or brands should we offer?”
- “Which other options did you consider before choosing our shop, and why did you decide to use us?”
- “If you were in charge of our company, how would you persuade people like yourself to use us?”
- “If you were in charge of our company, how would you “spread the word” about us?”
Another way on coming up with ideas is from Tim Ash, he suggests these “scales” to look for ideas.
Anxiety vs Trust.
Indeed I would have all these things regardless and test their prominence. I’m not convinced that trust symbols like BBB and GeoTrust are worth it, but they’re worth a test.
Confusion vs Clarity.
By emphasizing too many things we destroy visitors ability to find key information. This paralyzes them from making a decision. The best way to test this is through things like the 5 sec test or Clicktale.
Alienation vs Affinity.
Your tone and manner makes them feel understood and valued, part of my “tribe”.
For example I follow two expert SEO firms: SEOmoz and Bruce Clay Inc.I like the way Rand Fish in SEOmoz approaches his readership – friendly, slightly tongue in cheek, self deprecating and humorous in a geeky way. Bruce Clay Inc, one of their biggest competitors, on the other hand is more formal and professional. I identify much more with the Rand Fishes of the world than the Bruce Clays, so I ensure that comes across in how I write (ideally I should also be smiling in my photos but every time I try I look like I’m grimacing – very frustrating).
It’s not just single pages
The sales funnel is a multi page process: from landing page to cart to checkout to order to follow up to repeat business.
An analogy is a rickety rope bridge with missing pieces (from Tim Ash). The aim is to fill in the missing pieces to ensure that the customer is satisfied as they go from one end of the bridge, the landing page, to the other, the order.
You fix one plank at a time, then turn it into a concrete footbridge, then a highway, then … okay I’m getting carried away with the analogy ;).
Alcohol beverage marketers say that brand building is all about the right people drinking the right products at the right places. Tim Ash has a variant he calls the matrix where he says you ensure the right people go through the right activities in the right order.
The normal practice
For eCommerce fixing the pages closer to the checkout, including the checkout itself, tend to have a greater effect on increased sales compared to say the home page.
So test variants of:
- Headlines, probably sub headings on the Product Page
- Image – clear and present
- Product description copy – e.g. taste, terrior, people, reviews, ratings, humorous, conservative, shorter, longer
- Action block – with all the key info highlighted, price, quantity, add to cart button. Perhaps do a simple 1-2-3 step to show just what a person needs to do buy.
- The shopping cart – upfront shipping, guest checkout, single page checkout
I would just start with one experiment of the sub headlines on your product page (the headline probably being the actual Product name). As you start to get a feel for it try a few more experiments until you’re testing all the high traffic landing pages on your website.
And I might as well make the point here, some eCommerce software is very inflexible and you will be unable to do much testing. Might be a good idea to see what your web developer says. For instance Tim Ash (Aug-09) doesn’t have good things to say about Yahoo Store. Whereas I’ve had great experiences with Magento eCommerce and Google Website Optimizer.
A great article if you want to read more
The best practical article I’ve seen on this is Google Website Optimizer 101 – a quick-start guide to conversion rate optimization. The founder of this firm is actually an ex-rocket scientist – I kid you not.
In the next post I’ll look at installing Google Website Optimizer, what ecommerce software has this integrated (or not), a feel for costs and time, and a bit of statistical violence called “Beat the Control”.
Any thoughts so far?