Hey all. I tried a little experiment, and I am wondering if anyone else has tried the same.

When my games install, I always just put one icon on the desktop, and the rest in the start menu, just like everyone else. Well, despite it being slightly intrusive, I decided to place a 2nd icon on the desktop titled "More Great Retro64 Games".

My thinking was that this would generate more downloads and sales, because it would increase traffic to my site. Well, it did increase traffic. In fact, it almost doubled it.

But it seems to have a terrible side-effect. Sales sharply declined. I am getting the same amount of page-views on my order forms, but I am not converting nearly so many into sales.

I have a hypothesis as to why this would happen, but I am not going to mention it here- I don't want to pollute anyone else's perception.

What do you think?

heh you've just AOL'd yourself. AOL is always putting links to their free services on your desktop when you install various applications. I think people have a perception that when there's links to webpages on their desktop they get stuff for free (does anyone actually pay for AOL?), so when they get to your order form and see you're charging, they bail.

This is my theory.

Someone downloads the game, tries it, and likes it. But then, rather than go on and buy it, they decide to try out the other games to see if they are also this good. Boom! All of a sudden the thought of buying the first game is no longer at the front of their mind.

Similarly, if they are in the mindset to buy the game (getting as far as the order form), you are now making them think "well, if I'm going to buy a game, I want to get the best game on offer" so instead of buying the first game, they go off and try the others!

Lastly (and in my opinion the reason why download.com provides so many downloads) if you are telling the person you have a whole load of games that they can play, they can download demo after demo and get lots of play time without ever spending a dime!

That's my opinion anyway :)

Dax: your opinion is the same as mine... I only have 4 games, so perhaps after some amount of time, sales will go up. I have been doing it for 9 days, and I'll leave it up another 7-10 days to see if it picks up :)

Hmm very insightful. That makes a lot intuative sense to me.

If that is true, then maybe you should only create the second icon with the registered versions.

Yep Dax I think nailed it.

Let be ramble for a bit and see if any useful data comes of it....

We offer a combo pack, but when we started pushing it extra hard on the order page, we actually saw a drop in sales that week.

It might have been a combination of factors, but my take was the same...

Here's the deal as I see it. There are times when the more options I find out about the LESS likely I am to buy. I'm already wondering if there is something better to spend the money on like say food, then I see 5 cell phone plans, and they don't compare apples to apples. I do that to myself often, could be Cable vs. Satellite.

basically....

a) People don't like to part with their money

b) if they don't impulse buy (that is buy when you have the passion "I'm gonna get this cool game cuz is just rocks!", they can pretty easily talk themselves out of the purchase if they get distracted.

It might kinda be like "getting in the mood", Don't bring up taxes or what you saw on slash dot when you are trying to have a romantic time with your sweetie :)

One real world example, the store. Often time their will be some last minute options for you. They however are often complementary choices... "would you like a 2 year service plan to go with that printer cable" or " Would you like a something to drink with your order"

But when was the last time you ordered a Western Bacon cheese burger and they said. " Are you sure you don't want to get a Chicken burger too?"

Course most of this mainly applies to folks after they are on thier way to purchase.

In the case of the icon though, you are offering choices BEFORE they have the mind set to purchase....seems like a good thing.
" If you don't like this, try that".....better to get them now then to have to go out and find them again.

Let us know how the test continues.

Responding to Jonas' points... we've seen some interesting results in this area, which I mentioned in a previous thread. In our demo versions, the link to the order form is tagged with the name of the game being (potentially) ordered. So I tried an experiment where the game being ordered was the only one shown on the order form vs. listing all the products. This only took about 10 minutes to setup, since the order form is automatically generated from the database.

Sales were better when just that one product was the only one listed on the order form -- EXCEPT when we began offering a 20% discount for multiple purchases. In that case, sales were much better when every game was listed on the order form. Close to 50% of our sales for the past several weeks (during the time we've been offering this discount) have been for multiple games. Before the discount it was closer to 10-20%. And many of these sales are for 3, 4, ..., even 8 or more games purchased at the same time. We see individual orders for 3+ games every day now.

Another thing that seemed to help is that when someone clicks a link to order a particular game, we show the name of that game at the top of the pre-ordering page instead of just a generic heading. For instance, look at the headline at the top of each page when you click each of these links:
1) http://www.dexterity.com/buy.htm
2) http://www.dexterity.com/buy.htm?game=dweep
3) http://www.dexterity.com/buy.htm?game=pharaohscurse

I plan for us to do a lot more experimenting in this area, since its impact on sales is potentially tremendous. What discount rate is the best? If you list multiple products, is it best to list every game or just a cluster of games that are most similar to the product being ordered? Will some particular games do best if presented solo, while others are more likely to sell well in bundles? Is it better to present multiple games at the beginning of the ordering process, or would it be better to attempt an upsell after the initial purchase has already been committed?

Yet another option is to tailor this process individually based on the particular player. For instance, the order form could be personalized based on the visitor's previous purchasing history, if known.

This morning I've been working a great deal on our long-term strategic plan, and these types of optimizations will be a huge part of our focus in moving ahead. There is just too much uncertainty to be able to guess what approach works best, so we're just going to test and measure to find out what really works for us. That's the only way to be certain we aren't leaving loads of extra money on the table.

I have done TON of experimenting with my order pages. I've read several marketing books, copied direct mail ads, copied other sites, and tried my own layouts/copy.

And after all of that, I don't have a friggin clue what works best! The problem is that I would have to leave a particular method up and running for quite a while before it would be statistically valid- and I am chicken! :-) For some reason, I make a major change and expect sales to double for a product that same day. And when they don't, I change the page back to the way it was.

With so many multi-million dollar+ websites out there, I am AMAZED that there is practically no literature out there on the marketing and psychology of the web order form.

That basically leave it up to us to experiment and take what works best in our individual case. But I am sure we could all optimize the process and GREATLY improve sales conversions from order pages.

EDIT: Another question with the multiple product form is the actual order of the products. Is it best to put the game that the user came to buy FIRST on the form? Or is it better to place it in the middle. I also automatically check that game on the form for the user, maybe it's better to not check it? Who knows- only one way to find out I guess.

Recently I was pointing people to a list of all games. I then switched it to a list of the particular game + relevent bundles. So in the case of aargon you have...

Aargon
Aargon Gold
Aargon + level pack + hint guide

Sales of the 3rd much more expensive bundle are now the most common purchase whereas before it was aargon gold so profit is up from that change. The reason in this case is probably pretty simple. People just weren't looking for the bundle in the list. When it was made more clear it turns out that's what people wanted.

For the moment I've removed relevant bundles of different games because I'm worried that someone will stop buying to check out the other game's demo and then not come back.

I like steve's 20% sale. I think that's a great approach and perhaps that's similar to bundles of two dissimilar games. You should probably try adding 3 games for 30% off. I'd like to experiment with somethign similar but it's a little tricky to set up. I guess I'll have to bite the bullet soon and set up a merchant account which will allow more experimentation.

At the beginning of the year, we offered a $39.95 bundle of Dweep Gold and Fitznik, and that sold pretty well. But when we started adding more games, players started asking for different types of bundles. So instead of going crazy with too many bundles, we opted for the more flexible approach of a multi-game discount. It's been working nicely, since players will bundle things I wouldn't have expected. We just got another 8-game sale about an hour ago.

There are definite advantages to having your own merchant account, not the least of which is that you can control the whole process. Sometimes I shudder when looking at some of the registration services' default order forms, since they often include links that may distract people from ordering, such as a link to a compilation CD. If anyone should be committed to experimentation and optimization, it should be the registration services.

Of course, lower fees are another advantage to your own merchant account -- you lose only 2-4% of each transaction instead of 10% or more. A merchant account is more setup work initially, but I think it's the best option for anyone making more than $1000 per month in sales. You also get your money earlier, typically within about 1-2 business days.

As much as I think I know about selling games online, I'm fully aware that what I don't know is a huge area still waiting to be explored. For me that's exciting, since in that unexplored area, there could be some hidden gold mines of opportunity. There's always the potential to stumble upon a minor change that may double our sales permanently. I guess the important thing is to maintain that sense of awareness of what remains uncharted territory.

I disagree about having a merchant account. I really don't want to have to deal with chargebacks and fraud. Most registration services have worked very hard to auto-detect bogus orders in real time.

In addition, many registration services offer telephone ordering too. All this for only 10%, just a 5% difference from what you end up paying with most merchant accounts.

My registration service lets me configure my order form however I want. And I just now realized how bad that form is! It will be changed tomorrow.

As a 1-person company, I really don't want to think about those things, and my registration service allows me to not have to. Perhaps other services aren't as human or flexible as mine (BMT Micro), which may explain the differing opinions...

Another note about the bundle/multiple sales scheme. Here's something I've wanted to try for a long time. I'd like to send a time limited coupon in with my orders so customers can come back and buy another game after. So for instance...they purchase Aargon and then in their registration letter it says something like...

BONUS COUPON : Get 25% off if you buy another game in the next 10 days! Enter this coupon code "CUP102948" on the order form for a 25% discount.

Has anyone tried something like that? I think it might be effective if you could automate it. Some difficulty in automating the process is the only reason I haven't tried it yet.

Chargebacks and fraud don't require much effort at all, since in most cases, you just get a notice that a chargeback has occurred, so you don't have to do anything at all other than perhaps updating your database to reflect it. CardService International, our merchant account provider, also has a large fraud database, and you can also add your own blocks to the system. So the fraud-screening done by reg services isn't unique to them.

As for doing a coupon code, that would be pretty easy to do with an online database. Just generate a code that could be connected to a particular customer ID, and then when that code is entered on another order, just lookup the customer to see if they purchased a game in the last 10 days. In fact, it wouldn't be hard to generate a variety of different codes/offers (i.e. 5% off in the next 15 days, 15% off in the next 14 days, etc) to test which works best. This would be an interesting experiment.

Most reg services do much better fraud protection than a fraud database can. For example, they compare the user's ip address with the country they put in the order form. That alone wipes out many fraudulent orders. They perform other checks too, above and beyond what cardservice and other merchant account providers do.

Another advantage is the amount of different credit cards accepted. For example, my provider accepts Visa, Mastercard, Discover, Amex, Diners Club/Carte Blanche, Eurocard, and JCB. They also accept German wire transfers, which are surprisingly common. Germany is the 2nd biggest pc gaming market in the world.

This, combined with mail ordering, phone ordering, and less work for me, make it the right choice for me. I understand, of course, that different people have different needs. So I guess it's up to the individual.

How did a thread about icons on the desktop morph into this anyway? LOL :)

Actually I've heard reports of problems that reg services can sometimes go too far in trying to prevent fraud, causing legitimate orders to be lost. For instance, we get legitimate orders all the time where the IP address and countries don't match. We also accept orders that use free email addresses like hotmail accounts, whereas some registration services block these. Our fraud rate is way below 1%, so fraud hasn't been a big concern for us, although I do know that some businesses really suffer from it. Our fraud rate dropped quite a bit since we began offering a money-back guarantee.

The big advantage to using a registration service as I see it is what you mentioned in the rest of your post -- being able to accept a wide variety of payment options.

Originally posted by Dexterity
Actually I've heard reports of problems that reg services can sometimes go too far in trying to prevent fraud, causing legitimate orders to be lost.

This happens to me a lot. I'd say, without exaggeration, that I lose 10-40 orders a month due to strict fraud checks. Compared to actual fraud attempts against me, (very few as I can see them logged) this is rather annoying. Many reg services will relax the strict policies if your account doesn't have a lot of fraud. I imagine some products are more apt to attract theft than others.

As an expat I have a terribly hard time buying anything online. Usually what's supposed to be a 10 minute order is a 2 day order for me because my card is always refused and then I have to write support and explain that while my mailing address is in Canada I currently live in southeast asia therefor the ip doesn't correspond and so on and it's usually processed. Of course I'm "In the know" - must customers won't bother and the sale will be lost. I'm particularly appalled that most of the major vendors who implement these checks (as well as other strict checks like verifying that the address you used last time is the same as the one you're using now ... as if people never move) don't have a database of "good cards" for customers that have bought several products. I have to undergo the same tedious double checks and support emails every single time I buy a product on some sites.

Geez thats crazy that they are turning away 10-40 orders.

We probably have few a month.

My goal is simply to keep our fraud checks in line with our current reality. At present we average about 1 chargeback for every 500 orders, and this hasn't changed much in the past couple years. Even with the chargeback fees, it doesn't make much sense for us to put a lot of effort into reducing fraud. If/when it begins to show signs of a threatening increase, I would certainly look into tightening our restrictions.

I know that some developers block orders from places where fraud is reportedly high, especially countries in Southeast Asia. We see plenty of legitimate orders from Taiwan, Singapore, Hong Kong, and Malaysia without excessive fraud from those regions. Most of our fraud comes from the USA.