Want to use Pay-per-click to "split-test" everything - initial ad, main page content, price, etc

Hi... I'm close to releasing a new software I've programmed and developed, and maybe I'm asking for the moon here, but is there some slick way of "split testing" most (...or all) of the pertinent variables of my site/product? In other words, how to find the optimal pay-per-click ad that pulls the best; how to find the most effective content for my main web page to ensure maximum downloads; how to find the best price to maximize revenue.
   Just googling around a bit I see there's this "Google Content Experiments" that rotates pay-per-click ads randomly and shows you which one was most effective - that's pretty cool. Are there any more tricks of the trade or advice that could help me in what I want to achieve?

Thanks!
   Shawn

P.S: I'm a programmer tech guy... don't have much marketing know-how or experience.
shawn857Asked:
Who is Participating?

[Product update] Infrastructure Analysis Tool is now available with Business Accounts.Learn More

x
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

shawn857Author Commented:
anyone please....?

Thanks
    Shawn
0
shawn857Author Commented:
anyone who can help me??

Thanks
   Shawn
0
Lucas BishopClick TrackerCommented:
Yes, you can use Adwords to test a variety of components of your web site.  I'd recommend also having Google Analytics installed so you can measure engagement (time on site, bounce rate, etc.), Optimizely as it will allow you to A/B/C test a variety of aspects (copy, layout, CTAs, etc.) of your landing page and CrazyEgg so you can have a decent click map of your landing pages and identify of certain aspects are being ignored.

You'll want to integrate goal tracking in Adwords and Analytics also so you can measure success metrics (leads, conversions, downloads, etc.). You can link your Analytics & Adwords accounts so that the data is shared between the two systems and is easier to analyze.

You'll want to lay out your goals before you setup your campaigns, since you may need to setup a different campaign to test the success of each goal. Once this basic framework is laid out, you can start building your Adwords campaign(s).

You'll be able to measure a variety of things via your Adwords campaigns. For example:

Keywords:
What keywords drive the most qualified visitors?
Are certain keywords responsible for better engagement?
Are certain keywords associated with completion of goals?
Should certain keywords be put on a negative list so they never show with your ad?

Ad Copy:
Does specific ad copy coincide with better/worse engagement/success metrics?
Does specific ad copy influence a better CTR on your ads?

Landing Pages:
Do certain pages have better/worse engagement/success?

This is a very high level overview, but to answer your question, you can measure all of these things and more via Adwords. The real trick though is making sure you have your Analytics system properly configured so you can analyze what effects different changes to your site/copy/keywords/ads/etc lead to success for you.

You could use the "Adwords Experiments" feature which automates the experimentation process, but since you are just starting out, I'd recommend experimenting manually on your own at first. Once you've wrapped your head around the different things you can experiment with in Adwords (New keywords, New ad text, New ad groups, Negative keywords at the ad group level, Most keyword match types, Ad group default bids, including max CPC, Max. CPA, Keyword insertion) only then would I consider configuring an automated Experiment.

You could for example, setup an Optimizely experiment  with two (or more) different landing pages for the same offer/product. Each page could have different messaging or pricing for example. This way Optimizely can start to give you feedback on what works best for your visitors. I'd recommend searching out A/B testing best practices so you don't overdo it.

Within Adwords you'd simultaneously be testing out keywords/ad copy to drive visits to that Optimizely landing page. With a very basic test like this, you could start to identify the keywords/messages/offers that work best together.
0

Experts Exchange Solution brought to you by

Your issues matter to us.

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

Start your 7-day free trial
Exploring ASP.NET Core: Fundamentals

Learn to build web apps and services, IoT apps, and mobile backends by covering the fundamentals of ASP.NET Core and  exploring the core foundations for app libraries.

shawn857Author Commented:
Excellent info Lucas, thank you! Lucas, would you have any tips/techniques regarding how to test multiple product prices concurrently... preferably where Customer 'A' would not be aware of the different prices that Customers 'B' and 'C' are getting? That's a tricky area... I know. I'm thinking I could set up multiple copies of my website using different subdomains, something like:

subd1.mysite.com
subd2.mysite.com
subd3.mysite.com
etc etc

... each being identical with the only difference being the price offered on the Order page. Logistically though, it sounds like it could be a lot of work setting up and keeping track of all that...

Thanks
    Shawn
0
Lucas BishopClick TrackerCommented:
This is somewhat dependant on your ecommerce system.

Are you able to have the same product with multiple prices? Do you have to setup hidden products? Are you able to simply create a buy-now link with any price you want on it?

Assuming the eCommerce system isn't an issue and you can create buy-now links with unique pricing as needed, I'd do this via Optimizely. You'd simply create an experiment where the control is your current price and you setup variants for each price you want to test. On a single landing page you can edit the html/code of the page for each experiment, to adjust the price/buy-now link/etc.

Optimizely uses a cookie to put a visitor into an experiment version, so they would only see one pricing version unless they clear their cookies and come back to the same page.
0
shawn857Author Commented:
Yup, no problem with my eCommerce system... I'm pretty sure I could create multiple order pages for the one software product.
   I'm encouraged to hear I could do this with Optimizely! I've never really used "landing pages" before... so you are saying that I would just need only one of these, and from there the visitor would be re-directed to my main site index.html page, or directed right to my Order page?

Thanks!
    Shawn
0
Lucas BishopClick TrackerCommented:
In general when you market a single product/service online, a best practice is to use a landing page. This is a page whose sole purpose is to help accomplish a goal. You strip away all the distractors (navigation, external links, etc) that are normally a part of a web page that can cause visitors to lose focus and abandon your funnel. This way you have a controlled environment for experimenting.

You'd have a single landing page that is either built to capture the person as a lead that you can market to (ie. email signup form) or as a sales/marketing page that the person can buy from (ie. 1 click into checkout).

However, you can use your main web site (index.htm) or product page to run your tests on, its just that those pages need to be extremely well formulated (which is rarely the case) in regards to goal execution.
0
shawn857Author Commented:
OK Lucas, I like the idea of a landing page with a form to capture visitor info. I'm just trying to think how I'd dovetail this into my plan. My original overall plan of attack for capturing subscribers was to have them fill out my form and then immediately my autoresponder would email them with the download URL for my demo. The demo would expire after 21 days/uses (I can easily change the length of this). Within the demo I have a BUY button which would take the user to my site's Order page. However, if I utilize the landing page strategy with Optimizely, I'm thinking I can't have this BUY button within my demo, because once the user clicks that BUY button, my demo has no way of knowing which price or order page to use. See what I mean? So I guess to employ the landing page/split testing strategy, I will need to take out the BUY button from my demo, no? But in that case, how does the user who just completed my demo and wants to buy, get to his proper order page? At that point, I assume the user would need to be reminded to go back to the original landing page?

This is the sticky issue right now that I'm a little foggy on...

Thanks
    Shawn
0
Lucas BishopClick TrackerCommented:
Is the demo web based, or is it an app that is installed locally by the user? In either case, I believe this would be the process:

1.) You setup a unique form for each price tier
2.) You setup your optimizely experiment to show the proper form, for each price you test, this way visitors are only shown the form associated with their pricing tier
3.) You setup a unique autoresponder for each form/price tier
4.) The autoresponder provides a link to the demo version with a buy-now pricing tier that matches what the user was shown

So if you have 3 prices you want to test, you setup a single page experiment. On the single page the experiment features are the pricing (message) and the form (form id).
0
shawn857Author Commented:
Hi Lucas... my app is a Windows program that is installed locally by the user. That's the wrench in this whole thing I think - if my product was something like an e-book or online product that doesn't have a demo download, then all this testing and tracking stuff would be so much easier. Thanks for your outlined suggestion, but that doesn't really seem that I'd be saving myself much work with Optimizely. If I wanted to test say, 20 different prices simultaneously, I'd have to create and maintain 20 different subscription forms, 20 separate autoresponders, and the most work of all - create 20 separate demo EXE files.
   (I don't quite understand what you meant by the following: "On the single page the experiment features are the pricing (message) and the form (form id). ")

Before I posted my question, I was wracking my brain about this and came up with my own approach - which now actually seems like less work. I had hoped an advanced tracking/analytics system like Optimizely would make things far easier, but maybe it's just not possible with the way my software business is structured. Anyway, here is the approach I had basically cooked up: I would build only the one demo version of my app, but make several different identical copies of it, only assigning a slightly different filename to each. Something like this:

MyAppDemo_xyz.exe
MyAppDemo_abc.exe
MyAppDemo_rst.exe
etc, etc

These filename differences would mean nothing to the demo user, but would be my means of tracking what demo user is going to get what price. I would then create several "subdomains" off my main domain - matching the "xyz", "abc", "rst" found in my demo file name - they'd all be identical clones of my main website... the only difference being the price on the Order page. As a way to 'automate' this all somewhat and make less work for me, I was hoping it would follow this sort of process:

(1) Via pay-per-click, I direct my user to one of these subdomain sites ... let's say "xyz.MyAppSite.com"
(2) There he reads about my software, and requests the demo download site from me through my autoresponder.
(3) My autoresponder is configured to detect what subdomain-site the request came from, and emails back the user with the proper demo exe file. In this case, the user would get a response back saying the download exe file is called "MyAppDemo_xyz.exe" (my autoresponder would automatically grab the "xyz" from the subdomain passed and stored in this User's info, and tack that onto the demo filename as a 'custom token').
(4) User installs the demo and plays with it. If he decides he wants to buy, there is a BUY button within the demo that he clicks on. But therein lies the sticky part to this approach, where I would somehow need to find out what the full name of the demo file was that I sent him  (ie. MyAppDemo_xyz.exe). If I could get that info, or track it somehow through the whole process, then it would be a simple job for my demo to grab the last 3 characters of that demo file name, and direct Mr. User to the proper order page when he clicked my BUY button.
 (ie. xyz.MyAppSite.com/Order.htm)


Elaborate and convoluted, perhaps... but it's the only thing I can up come up to test multiple prices in some sort of semi-automated setting where I wouldn't have to be manually creating and keeping track of a million things.
   Surely I'm not the first software developer in a similar situation who has come across this problem. I'm thinking that somebody must have devised a procedure or method of some sort on how to do effective and easy price testing for someone in my situation. Seems to me I can't be the first.

Thanks  
    Shawn
0
Lucas BishopClick TrackerCommented:
If I wanted to test say, 20 different prices simultaneously,

A/B/C/D/.../T testing would require an enormous amount of users to provide any statistically significant results, especially considering you'd be testing 'dollars per visitor' or 'dollars per demo'. In most cases I'd say testing 20 [things] against each other is simply too many [things].

I'd have to create and maintain 20 different subscription forms, 20 separate autoresponders,
Programmatically, you could use 1 subscription form, with a hidden field that contains the experiment price (which is delivered to the auto-responder). The auto-responder would provide the appropriate download link in turn, based off the hidden field.

In this scenario you'd need to create a demo .exe for each price configuration (I'm not sure if that is difficult on your side.. it sounds like creating 20 different versions would be). I don't know enough about app dev to provide guidance here. In your scenario it seems like if the user changes the .exe file name, you'd potentially have an issue.

(I don't quite understand what you meant by the following: "On the single page the experiment features are the pricing (message) and the form (form id). ")

Basically on the landing page you'd only be swapping out the price (if you advertised price on the page) and the form id or hidden field (posted during the form submit), for each experiment.

With my limited programming knowledge I can only see creating a demo.exe for each price test. Maybe there would be some sort of price.ini file included with each app version that reduces your overhead? Maybe the app can be compiled on the fly upon receiving some data from the form submission? I'm probably extremely off-target here, app dev is outside of my wheelhouse.

The issue you need to resolve is on the app dev side. How to create a single version of a windows application, that can have dynamically updated buy-now price references, based on a price parameter that is not provided within program. This is probably a good question to post into the appropriate programming language sub-category here, as it's very far outside the realm of my area of expertise (internet marketing).

In any event, the A/B test process is fairly straight forward on the PPC side. You can identify the ad copy, keywords, web page copy, advertised pricing, etc. all through PPC.
0
shawn857Author Commented:
Yes you're quite right Lucas - testing 20 prices simultaneously is just not realistic... even 4 or 5 at a time is a stretch, but that's what I'm going to try and we'll see what happens. To create 4 or 5 different separate demo EXE files is not a huge deal and I can keep track of all that pretty easily.
   Lucas, thanks very much for all your help and advice - it's been invaluable.

Cheers
    Shawn
0
shawn857Author Commented:
Excellent.
0
It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
Internet Marketing

From novice to tech pro — start learning today.