“It’s a numbers game”
This well-meaning advice that we’ve all heard time and time again has the ability to cripple our website’s profitability.
Yes, it’s true in scripted sales that if you have a product that you know sells and you deliver the same script over and over again then eventually someone will buy from you.
When you are told “It’s a numbers game” you are being advised to get through as many presentations as quickly as possible so that you make the most transactions. It’s simple math, if you knew that a certain percentage of your presentations would result in a sale then you’d want to make as many presentations as possible. Many a cold-calling sales professional has been sold on this premise. This is the same logic that results in website owners doggedly focusing on traffic; any traffic will do – “it’s a numbers game”.
While this idea of a numbers game may help to reduce anxiety of cold-calling and rejection in individual salespeople, and may work in the short-term if you have a lot of leads or traffic you can burn through, it is absolutely crippling to your online business.
Why is that?
Well let’s imagine we hire two telemarketing salespeople, Joe and Joanna. Joe closes on average 1% of all his scripted telephone presentations, and after a day calling 200 prospects he has made two sales. This is an average, ordinary day for Joe.
Joanna has customised her script a little in line with her personality and she uses a really friendly, jovial tone of voice over the phone. She closes an average of 3% of all her scripted telephone presentations. After a day on the phone talking to 200 prospects Joanna has made six sales. This is an average, ordinary day for Joanna.
Both of these salespeople are playing a “numbers game” in that they can be pretty confident that even if they hear a lot of ‘no’s in their day they will still hit their averages, and so they just keep going until they hear a yes. From the perspective of the salesperson it is a numbers game.
But as a business owner, if you are hiring Joe and Joanna, you will be very aware of the fact that Joanna is making six sales whereas Joe is only making two, from the same number of leads.
From the perspective of the business owner, this isn’t a numbers game, this is a situation where the business owner wants more people like Joanna and less people like Joe. Because Joanna brings in three times more business for the same salary, she is much more valuable than Joe. The business owner would most likely prefer to have two Joannas in the sales roles and to find Joe a job elsewhere in the company.
Now imagine that Joe and Joanna were webpages on your website. On average, when a visitor looks at Webpage Joe, a sale is made 1% of the time. On average, when a visitor looks at Webpage Joanna, a sale is made 3% of the time. Now wouldn’t you want all of your website visitors to look at Webpage Joanna? Wouldn’t you want more webpages like Joanna, not like Joe?
Well guess what?
Your website is riddled with Joes and Joannas but unless you invest in a little analysis of your numbers, you won’t know which pages are which.
Here is an example for you. Imagine a business that sells furniture and this furniture is great quality but should never be kept outside. Imagine that this business has a page dedicated to explaining why this furniture should never be kept outside. No products are sold on this page.
Doing a strategic analysis of this business’ website I discover that most visitors who viewed this webpage left the website without buying anything. Visitors who did purchase goods from this furniture retailer almost never visited this page. So this page is not a good sales page for the store. It is a Joe. There are other ways to share this information without giving it a dedicated page.
What you want to do with your website is weed out all your sales Joes and upgrade them to sales Joannas.
Here is what I would do if you asked me to find the best and worst sales pages on your site.
Firstly, I’d need to know what you are trying to achieve with your website. Is it direct sales, leads, donations, or some other type of engagement?
Next I’d set up a Goal in Google Analytics that gets triggered if a visitor achieves that purpose of your website. Let’s assume for now that this website has goods for sale and if the visitor buys any one of your goods then the Goal is triggered. (If you wanted to get really specific you might want to target a visitor who buys goods from you and also spends over a certain minimum value, but just setting the Goal as a sale is easier to implement and provides you with data sooner). To learn more about adding a Goal in Google Analytics I will publish a post about this shortly.
Next up I’d look at the Conversion Rate and traffic volume for each landing page that visitors are going to on your website. You’ll see that some pages have large numbers (even as high as 100%) and some have low numbers (perhaps as low as 0%). Many of these pages will have high traffic volumes and Conversion Rates that look like 1%, 3% etc, just like our Joes and Joannas.
A quick look over these numbers will give us an idea, but we don’t want to stop here. If we took these to be our final numbers we’d be in trouble.
Why is that?
Well, you’d find the thankyou page that your visitors go to after they purchase something from your website, that page would have a conversion rate of 100%. And the page that almost no-one ever sees, because it was published 15 years ago and you don’t link to it on your homepage anymore, well perhaps one person found it in the last month but didn’t buy anything. That page has a conversion rate of 0%. Neither of these pages are active sales pages.
Secondly, what we’ve done here is just look at the landing pages, but we haven’t looked at all the middle-of-the-sales-process pages which might be more relevant. Google Analytics doesn’t have a standard report with a Conversion Rate for all pages, it only tells you the Conversion Rate for landing pages. (If you want to know why, this is because Conversion Rate is a session metric and Google Analytics records sessions for landing pages and pageviews for all pages).
What you want to do is find the pages that were involved before a sale, and then compare how often they were viewed before a sale compared with how often they were viewed when no sale took place. To do this you need two segments – one for users who purchased from you and one for users who never purchased from you but who looked at more than one page. To learn more about adding segments in Google Analytics I will publish a post about this shortly.
(Note: You want to consider browsers in order to ignore visitors who just looked at one page and then left. Otherwise some pages might look like Joes when really they’re just high-bounce landing pages, a different beast altogether. Also, since some users might leave and then come back and buy a few days later, you want to use at least a month as your timeframe. A good rule of thumb is at least 100 purchases.)
Your Joannas will be the pages that were often viewed before a sale but were not often viewed when no sale took place. Your Joes will be the pages that were often viewed when there was no sale, but not often viewed when there was a sale. The bigger the difference the better or worse the page converts in general.
OK, so what does this look like in practice?
Let’s say you have one page selling sports shoes. You look at your Google Analytics and find out that this webpage was viewed before 9% of sales on your website. On the other hand, when people browsed your website but did not buy anything at all, they only looked at the page selling sports shoes 3% of the time. This page is a Joanna. As long as your assessment is based on enough volume of visitors then it is likely that if you directed more people to this page then you would have sold more shoes.
Now let’s consider a page selling ballet shoes. If only 1% of people who purchased a product looked at the ballet shoes range before buying, and 6% of the people who looked at the ballet shoes page did not buy from you, then this ballet shoes page is a Joe.
(You might think that you would already know this if you had sold lots of sports shoes but not many ballet shoes, but not all pages are so cut and dry on your website. You will have products that are sold on more than one page and this makes it hard to tell which pages are working, even more so if you’re measuring email leads and your opt-in is on a sidebar of every page of your website).
You will find that if a webpage has nothing at all to do with sales or is in a long tail of product options, it will have pageview percentages of 0-1% irrespective of whether the visitor is a purchaser or a non-purchaser. Ignore these for now.
If a webpage has everything to do with sales (like adding to cart, checking out and finishing order pages) then these pages will have a very high conversion rate in the purchaser group. Ignore these too. Better yet, filter these sales pages out using the advanced filter. This will update all the pageview percentages for the purchasers to make the numbers more accurate when comparing to non-purchasers.
You are looking for pages in which the pageview percentage is consistently higher in either the sales or non-sales segments. My rule of thumb is that you’re looking for the % of total pageviews to be double that of the other segment if the highest % total pageviews is less than 10%, or 5% difference if you’re working with larger numbers. The smaller the percentage, and the smaller the volume of traffic viewing the page, the the less accurate it is likely to be.
You might have a lot of pages on your website, so here is a tip for more quickly finding the effective and ineffective sales pages. Firstly, sort by the pageview column. You are looking for pages that have a large difference between the percentage they were viewed by purchasers and the percentage they were viewed by non-purchasers. These pages will naturally rise towards the top due to having higher volume and you will find them in the top 10-25 pages of your website depending upon which segment is listed first. Then, switch the order of the segments and check the top 10-25 pages again. The top pages will change depending on which segment is listed first. To save time, I use a script to do this but you can definitely do it manually, looking for view percentages that are quite different.
OK, so now that you’ve found your Joes and Joannas what do you do with them?
I recommend that you list out all your Joannas and all of your Joes and look at each of the pages to see if you can see a pattern as to why a particular page is being viewed relatively more or less before a sale. You might even want to print the pages out and stick them to the floor or a wall, Joes on one side and Joannas on the other. You might find that all of your worst pages have a common characteristic such as a problematic user interface, or relate to a topic that doesn’t appeal to your buyers. Perhaps the titles, copy or images on these pages are not as good. Brainstorm and come up with a list of ways that you can make your Joe pages more like your Joannas.
WARNING! Don’t just delete your ‘dud’ sales pages! One reason that a page may not perform very well is because it is bringing in a lot of organic traffic but then not converting very well as many of the visitors are not the right target market for your website (but some are). If you removed these pages altogether you might find that your overall traffic decreases, reducing the number of sales overall, even though the higher quality pages were remaining.
Unless you know a page is not a good landing page for traffic generation purposes, I suggest you keep a high-volume page intact but with changes to make it more like your best pages. Alternatively include more calls to action for the visitor to get off this Joe page and onto a Joanna page. Some high volume landing pages will always convert less, just because the traffic viewing those pages is of a lower quality.
I also recommend you split test any changes you make before you make your changes permanent. The reason being, your website is a complex interconnected system and you never know how a change might impact the system, even if it seems like an obvious fix.
If you are interesting in learning more about how you can find out which of your pages are doing a great job of selling, and which ones are just your average Joes (ha ha!) then please contact me!
If you would like assistance with analysing your sales pages to find out which are performing and which are not then please send an email to Petra Manos ( firstname.lastname@example.org ). Web Data Analytics loves to help businesses grow their conversions online.