Visit us on Google+ googlee60acfabd67878db.html

A Conversation with Ian Spencer on How To Approach Your Online Testing

AB Split testing

This month we interviewed Ian Spencer of IS Digital Marketing to get his insight into the world of online testing.  Ian is highly knowledgeable, as well as being a qualified Google Adwords Partner.

His insights are fascinating and I hope they help you when thinking about what marketing testing you can do for your business.

Could you explain to us what online testing is and its importance?

Online testing to sum up in one sentence is pretty tough, but generally, it is the process of testing two versions or two events to see which one gets the better results.

In practice, you might want to test which web page gets more sales, which AdWords advert leads to more clicks or which jumper you sell performs best on Facebook, which is where Online Testing comes in to play.

Most clients use online testing as a way to improve and tweak rather than make radical consumer based changes, but however you use online testing, just consider it a good way of working out the right sales path for your customers or potential customers.

What can you test online?

As I mention later, my core testing often falls into:
Split Test (or A / B Testing)
Campaign Testing
Website Landing Page Testing / Optimisation
But you can pretty much test anything.

You might want to test a product price point in the market, you might want to test two versions of your website to see which works, or which email newsletter format works best for your customers.

Generally, if you want to test two versions of something to see if it lands more sales, enquires or results, online testing is the way to do it, but of course, you will need a website or product placement online to test (stating the obvious!).

Who should carry out online testing?

Anyone with a website should be looking at testing.

Whether that is testing new landing pages for AdWords or looking to see if you can improve the sales performance of a product, the online world is often instant and quick to decide, so it is the one place you can try two different versions of something and really see just how successful they are and then make a decision about the best way forward.

When should you test and how often?

As often as you need to, but do not get wrapped up in the need to constantly test, as you have a business to run as well.

Is it easy to do?  Can mere “non techy” mortals carry out online testing for their business?

I would not say it is massively easy no.

You need to make sure you have a way of measuring your tests. So, if its conversions, have you got the right conversion tracking installed, if its more analytical content measurement then have you got analytics installed correctly and do you know what you are looking for in terms of a test result to base success or failure on?

Know what you are testing for is obviously important, but having the right tools and knowledge to analyse your test results and then either make tweaks or improvements going forward is pretty important as well.

For someone who’s never tested before what are three “must do’s” and three “must not’s” for online testing?

Must Do’s:
Fully understand what you are trying to test, whether that is more sales, more time on site, or any other reason, as unless you have a goal, the test if worthless.
Use analytics and fair reports to measure test results, otherwise your results are not going to be an accurate reflection and therefore, once again, the test ends up being pretty pointless.
Take your time to really monitor the results and give yourself realistic time frames to not only give the tests time to work, but to also get enough data to make sure you can base firm results from the tests you have carried out.
Must Not’s:
Do not test for the sake of testing, otherwise you are just wasting time. Work out what you want to test and why, and then decide whether it is worthwhile or not. Some tests just do not need doing, will cost money and will then just confuse you, so have a clear idea of what you need to understand and then see if online testing is the way to understand it.
Do not rush the results, every test needs a sound time frame and for you to stick with it, do not be tempted to close the test early just because you want quick results.
Don’t skimp and save because you do not want to invest the money, as you need to spend some money if you want to get quick and accurate results.

Most of us have our site hooked up to Google Analytics, but how can we use this for effective testing?

Again, this comes back to what you are trying to test.

But, you can use Google Analytics to measure many things, from conversions through to time on site, so it all comes back to what you are testing and how you are going to want to measure results, it hard sales, enquiries, shares or any of the various other things you might set as your goals.

Generally, Google Analytics can be used how you need it to, with the use of Goals for conversions and a wealth of data that once you know what you want to test, you can normally find a way of using it to see results.

What are the top three things you can test on your website?

For me, across all of my clients, I would say you can test:
Product Feedback
Content Measurement
Website Conversions
These are the top three things we look at using the online world to test, as we can quickly see how popular a product is and if it’s the correct price point, we can measure how successful our content is and we can also look at website conversions and test different landing pages.

What are the most common incorrect assumptions people make with online testing?

That you will be instantly able to tell what is right and what is not.

This is not the case.

You need a structure in place to know what you want to test, how you are going to test and then how you will analyse the results. Just because Campaign A delivered more sales than Campaign B does not always make Campaign A more successful, as did you spend more to get the results, were the product sales of lesser value, were the enquiries relevant etc – the point here is that a quick look at figures does not always indicate the correct test result.

So, unless you fully understand what you are testing and then you fairly compare both tests, the assumption that you will instantly know is just not correct or fair.

Can you share with us an interesting insight you’ve gained from a project that previously wasn’t been tested and then was tested?

I recently took over a large AdWords campaign for a company in London.

The basics were ok, but they were running one advert to one page on their website.

So, we made 4 new adverts and effectively doubled them, sent 4 to one page, 4 to another and we were then able to see if different pages did better than others. We then used both conversion tracking and onsite analytics to work out which kind of advert worked best for this clients audience and then made the rest of the campaign using the same tone and of course, landing page which converted the most.

This is just one example, but across the board, we carry out tests across many aspects like this every day.

If someone has never tested before, what’s the one thing that they can do today to start testing online?

Spend some time to work out what you want to test, how you are going to test it and then how you will measure the results.

Failure to do this just means you will be wasting your time. Testing is only worthwhile if you have something to test.

If you would like to speak to Ian further about online marketing for your business, he can be contacted via his website or on 0330 223 2270.

Leave a Reply

Your email address will not be published.

* Copy This Password *

* Type Or Paste Password Here *