How to beta launch an eCommerce site?

Beta launch processes

[maserati86]

Dear all,
We are looking to test a new eCommerce site versus its current version and before cutting over, we need to test all major pages (products, booking engine and the shopping cart) and see if they perform better than the previous version.

Here are the options we came up with:

1) Redirect a certain amount of global traffic coming from all sources to the new site while the rest land on the old site.
We are still debating which tool to use for that redirection and if such a redirection (javascript based) could hurt our organic rankings?
2) redirect traffic based on specific IP range. For instance all visitors from Colorado will only see the new site
3) Add a link to the current home page "visit the new site"
4) Use different destination URLs within an ad group on Google Adwords.

All options have their pros and cons, in our opinion only the option 1 is statistically correct but we are open to suggestions

Let us know your thoughts
Best regards

[Marshall]

Option one seems to make sense, but if it is a beta version and encounters problems, you may lose some customers. Have you considered placing a link on the current site and asking people to participate in a beta version test? Maybe offer an incentive like an extra 5% off for their feedback.

Marshall

[lexipixel]

Is it really "BETA"? Have you beat it to death in ALPHA / in-house testing ?

I like the #3 option -- a link on the established site to the new one.

Either way, I think it depends on the overall amount of traffic, the total number of products, product options, categories, payment and shipping methods, and if you handle multiple currencies, have assorted customer types, run any affiliate programs, need to test CRM ticketing system, promo codes, etc..

If it's not a small simple store, then I'd run both sites for some time -- leave the older established site up as www.example.com and run the new one as store2.example.com, (or some other domain or sub-domain). That makes it easy to temporarily close the new store if there is a critical problem.

I'd invite established customers to use the new store first, (ones you can count on for good feedback -- ones that know how the old store worked and can tell you what they do or don't like about the new one).

Even if you get all the products, prices, images, etc.. all looking good, you need to find out if order processing, inventory control and any CRM, marketing or reporting functions are working.. etc.. etc..

It can take a while to fine tune a large ecommerce site. Since money is involved, I'm sure you wouldn't want to find out you didn't test enough and are losing money, rankings, customers or anything else.

You asked, so that's my $0.02

[maserati86]

Thank you for your input to our Beta process reflexion. You are correct in pointing out that putting a link on the current site and/or invite good/past customers to experience the new site and gather their feedback is the way to go. The problem is that we don't have many returning customers on the BtC site because we are a destination AND transportation specific travel site and people book with us every 2 to 3 years at best.
Also we will start UAT next week, followed by a round of qualitative customer testing with a Go/NoGo at the end of this round.
I think the "alpha" period is taking care of but we are still not sure if we indentified all Beta options and some of their dependencies (like losing ranking on Google if we redirect traffic via JavaScript)
FB