Should I test my payment integrations before launching? If so, how?

[Ultra Magnus]

I have a WooCommerce store set up on my sales letter page. I installed integrations for PayPal and Braintree, the latter of which was a purchased plug in, while the former was just offered as a standard payment option in WooCommerce.

I went through the process of having my accounts vetted by the processing companies. Braintree was particularly thorough before giving me a Merchant account, but I passed. I also sent a lot of personal info to PayPal through my account there.

Through it all, I've seen the "sandbox" options in the WooCommerce Settings tab. I also sought advice from my friends who had previously set up a store, but unlike me they integrated all the payments themselves and used Presta, not WooCommerce. They suggested that I should test it with a cheap bogus product. In effect, I want to ask other entrepreneurs the following:

1. Is it necessary to test these two payment options before launching? I was thinking yes, but then I don't know what to test since I didn't program the integrations myself. I could test notifications, emails, and payment speed maybe?

2. If your answer to the above was yes, how should I go about it? As I mentioned, my vetting process with Braintree was very thorough. I have only one high value product, and adding a "test option" would require notifying them and asking if it's okay. Should I use "sandbox" mode instead?

3. What did you do with your store?

I would be very grateful for any responses from folks in the online trenches, especially if you aren't a programmer and went through a similar process.

Thanks in advance!

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Jon L]

yes...test it. I develop software. A feature that you just 'assume should work because a big company made it and it works everywhere else' is the feature that will not work when you go live.

If you change settings in your woocommerce store later on, test everything again.

Software is complex, and everything affects everything else. One section of the program "shouldn't" affect a seemingly unrelated section, but it sometimes does.

 

[Ultra Magnus]

Thank you, Jon. Do you have any suggestions for the best ways for tackling this? What do you usually do?

 

[Ryan Wolf]

Not sure if this is what you mean, but I used a few test credit card numbers prior to my launch to make sure everything was set up right. Thank god that I did, too.

Here's the link for that info: Testing

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Lex DeVille]

Thank you, Jon. Do you have any suggestions for the best ways for tackling this? What do you usually do?

If I were you I would read through PayPal's instructions on how to use Sandbox. There's plenty of info on Google about how to do it. Or just make a woocommerce product for .05 and checkout under a different account that doesn't have an email associated with your PayPal account. Not sure why you'd have to notify anyone about adding a test option.

What you need to test is whether or not your checkout system works. Can you checkout? Can money actually get through to your account? If not, that's a problem.

 

[Ultra Magnus]

Thanks a bunch guys!

@SinisterLex, the guy at Braintree who handled setting up my Merchant account told me to inform him about changing/adding products, since they accepted me based on what I showed them. The website is a long form sales letter with a few buy now buttons for one product, rather than a regular store with a lot of stuff available.

Braintree also has a sandbox account option, I could check that out alongside Paypal. My questions arise partly from the fact that Braintree processes credit cards directly to my bank account, so I can't really check if the money reaches me with a "dummy" option that doesn't involve real money.

 

[Ultra Magnus]

I just called Braintree tech assistance and they advised me to create a sandbox account. So that's what I'm gonna do with PayPal as well.

Here's a valuable lesson which I usually follow and didn't this time:

Don't post threads asking technical questions that a paid support team was created to answer. I very much appreciate the opinions posted here in terms of direction, but I understand that asking the "hows" was a bit of a waste of time for you.

It's not much, but please accept some forum buckaroos for your trouble.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Jon L]

Here's the super anal approach to testing:

Create a spreadsheet called a Test Matrix where you list out all the possible ways an idiot can use your website. (What if they add an item to your cart, then remove it, then add a couple more items, change the quantity, save it, close their browser, come back to it from a different browser on a different computer and select two forms of payment when checking out?) (ok, maybe not that drastic, but 'idiots' are very creative.)

At a minimum, you want a few test cases - single item order, multiple item/quantity, different payment types, etc. You certainly want to test the scenarios that most everyone will use on a regular basis.

Then: when you make structural changes to your site, run through the test matrix again to make sure everything is still behaving normally.

When you add features to your site, add a test for those features to the spreadsheet.

 

[Ultra Magnus]

Thanks a lot @Jon L!

Definitely a great method when I have more products in the future (I have just one now, and it's a limited edition).

For now I've tested a Visa payment in Braintree sandbox mode and it worked fine. I'll test all the other cards and then move on to PayPal.

 

[miked_d]

I definitely experienced issues when moving from PayPal's sandbox to live. As @Jon L said, test thoroughly in the sandbox and then retest everything live. You should be able to lower the product price during testing so that funds are not an issue.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[ravenspear]

I definitely experienced issues when moving from PayPal's sandbox to live. As @Jon L said, test thoroughly in the sandbox and then retest everything live. You should be able to lower the product price during testing so that funds are not an issue.

Why would funds be an issue even if he paid full price? I hope the money going in the merchant account is going to him lol.

Also not sure if the OP has full access to the credit card API (sounds like maybe not) but you can also reverse a charge right after you order before it settles.

 

[miked_d]

Why would funds be an issue even if he paid full price? I hope the money going in the merchant account is going to him lol.

Also not sure if the OP has full access to the credit card API (sounds like maybe not) but you can also reverse a charge right after you order before it settles.

I don't know the price of his product. Anything over a few bucks will quickly start moving money around once he finds a bug (he will). Moving funds from your PayPal account is not instant. Reversing payments is not instant.

I am talking from experience, having done this with two different PayPal API systems.

 

[Jon L]

I don't know the price of his product. Anything over a few bucks will quickly start moving money around once he finds a bug (he will). Moving funds from your PayPal account is not instant. Reversing payments is not instant.

I am talking from experience, having done this with two different PayPal API systems.

Paypal has millions of lines of code that run everything over there. There's absolutely no way that its all bug free. The chance that you'll run into a bug is pretty slim, but you're betting your company on it, so it makes sense to test.

Also, when you integrate an ecommerce system with PayPal, all the code in the ecommerce system (with all its bugs), and all the code in PayPal (with its bugs), has to work flawlessly together.

Add to all that the strong possibility that you checked the wrong check box somewhere at some point when you set up the site, and you have decent odds that there's a bug somewhere.

Stuff that is mission critical needs to get tested (and getting notified of an order and paid for it are pretty mission critical). If you don't, you're asking for trouble.

I used to work at a company whose software processed billions of dollars worth of transactions per year. At any given time, we had a list of known bugs several-hundred deep. Those were the ones we knew about. Every few years or so, a major bug would crop up and cause a major disaster somewhere. 100% guaranteed PayPal is the same...

Go test.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Ultra Magnus]

Well, the product I'm selling is 3 bucks shy of 300 euros, plus ca. 20 euros for shipping... I can't imagine testing it with real money, at least not until I sell some Also there are transaction fees, which at these prices and at a moderate volume would be a bit ridiculous just for testing.

@ravenspear
In my Braintree dashboard I can search for specific transactions and then get a receipt, clone the transaction or refund it. Perhaps if I keep the dashboard open while placing an order I could find it quickly and reverse it...

In terms of API I believe I have all the rights I can have (it's a separate login on a different site than my store back-end). The integration with the WooCommerce platform however is a commercial solution I just bought as a plugin – I have no idea how it works under the hood (if I did I'd probably code my own using the docs from Braintree).

Also, since I never clarified it, I have the Braintree integration specifically to accept credit/debit cards because PayPal forces me to convert money at unfavourable rates. I know it's also a PayPal company, but the transactions don't go through that system. It's like Stripe (which is unavailable where I live).

These are all very interesting comments, especially since I'm not a techie and don't know what I should expect in terms of functionality. One thing I've learned from setting up my page and another lead gen page with a downloadable pdf is that, no matter how childishly simple the graphic design in the marketing of a web-based product is, you still have to go through a lot of bullshit to make it work. Or, as a programmer friend of mine succinctly put it while helping me with a feature: "If it was that simple I wouldn't have a job".