• Set aside a marketing budget.  We allocate 10% profit per quarter to go to the marketing pot.  This could fund entering competitions, promotional material, and a personal goal is to eventually be able to attend a trade fair.
• Set aside marketing time.  Although most marketing methods detailed in part 2 are free, they can take a significant amount of time.  This is time that is not spent coding or designing or doing art, so there is still that opportunity cost to consider.
• If something seems too good to be true it probably is.  Be wary of emails from “game marketing consultants” offering to market your game.  Imagine our surprise when we got an email from a “company” who boasted Valve and Activision amongst their clients yet whose website consisted of a single page and a domain that had been registered the week before!  We’ll steer clear of that thanks.
• Don’t give away all your promo copies right away.  You only get 50 and when they are gone there are no more (although there is a clever way to find out if your codes have been used or not and re-use them).  We’ve missed out on a couple of chances by not being able to offer promo codes to what looks like great marketing opportunities.
• CafePress offer a good service.  It is expensive, but considering CafePress deal with everything from taking orders, creating the goods, packaging, shipping and dealing with money, it’s a good deal.  This allows your team to focus on making games rather than running a mail order business.
• Game Sales and Website Traffic are Unrelated.  Game sales don’t seem to drive website traffic, nor does website traffic seem to drive sales.  Perhaps it’s a mixed two way flow with a variable time lag.
• Leafleting is hard work.  To get into shops is hard, and to do it yourself is also hard work.  Even in a city like London it is slow going and a lot of walking to do this yourself, plus for cheap games like the Gerbil Physics series it isn’t financially viable.
• Players Ignore “Other Games” Screens.  Suggesting players try another title on an “Other Games” screen in-game had zero measureable impact on trials of the other game.  Either the game didn’t look attractive or nobody pays attention.  Either is possible.
• The internet contains hate.  No surprises here, but no matter how great your game might be, someone will hate it and will be delighted to share their views (without any prompting needed) in comments on forums etc.
• Re-entering the Xbox Charts is possible.  It is worth reiterating the point that at present, you don’t need to shift an insane amount of copies to make it into the top 20 downloads.  It seems to me easily possible that with a concerted marketing effort, all focussed on the same time period, that any game could make it back in to the top 20 downloads and so get a new lease of life.

We’ve just started the main marketing effort for our latest game, Gerbil Physics 2, which has just been released.  Let’s see if we can practice what we preach .  If anyone has any great marketing ideas, let us know!

The Results

So, some hard facts then.  Our analysis period is 7^th Dec 2009 to 30^th April 2010 during which time we got 21,115 trials and 7,208 sales.  This gives a reasonably healthy conversion rate of about 34%:

Chart of Gerbil Physics 2 Sales

The sales chart shows 3 distinct phases of the game’s life so far.  Phase 1 was the first week or two after release, when the game was on the Top Downloads and New Releases charts on the dashboard.  This faded quite quickly but we were fortunate to then be featured by the IGN Top Picks page on the dashboard which was Phase 2.  

This lifted sales again for another few weeks.  Once we were cruelly removed from the IGN page the game reaches Phase 3, which runs from early January to today, where the game is bumbling along steadily selling 30 to 40 per day.  Most of these sales we believe are due to us still being in the Top 20 Best Rated chart in the US and UK.  Phase 3 shows regular spikes in trials and sales where we sell more over weekends than we do on Mondays and Tuesdays.  

We anticipate an additional Phase 4 when we drop off the Top 20 Best Rated.  Then we would then expect to sell only a handful of copies per day, and would then be relying on word of mouth and, well, marketing.  

It is interesting to see the graph of the game’s website hits over the same period:  

Chart of Gerbil Physics Website Visitors

We don’t get a wild amount of visitors as you can see.  Visually, there doesn’t seem to be any correlation between the sales and website graphs, but our website traffic is so tiny it is easy to see when the press releases happened.  The Valentine’s Day press release is clearly visible.     

Although there is no visual correlation between sales and website hits, I did a fair amount of work trying to find a numerical correlation.  I thought perhaps game sales could drive website traffic, and there would be a time delay between people buying the game and looking at the site.  Or perhaps website traffic might drive sales, and there would be a time delay between people visiting the site and then buying the game.  My analysis of the data shows that neither is true, and all scenarios I could come up with showed sales and website hits to be almost perfectly uncorrelated!     

The Costs and Our Subjective Opinions on Benefits     

Next then, here are the costs (approximate US Dollar figures) for each of the marketing strategies outlined in part 1, together with our subjective view on their benefit:     

• Xbox Marketplace Charts  |  Cost: $0  |  Benefit: High
  The Xbox Live dashboard charts gave us the biggest exposure, allowing over 1,000 trials to be downloaded on one happy day just after launch.  Being picked up by the IGN Top Picks was also a big boost.  One really notable fact is that, at least at the moment, it doesn’t take that many sales to sneak in to the bottom of the Top 20 Downloads.  If you could sell maybe 50 copies in a day (this number will vary by region) you could enter the top downloads chart and suddenly your game’s visibility shoots up.
• Press Releases  |  Cost: $0  |  Benefit: High
  We’ve estimated the benefit of press releases as high because it does drive website traffic and is a great way of getting some general exposure, even though we’ve no numerical evidence to correlate press releases with sales.
• Free Games  |  Cost: $0  |  Benefit: High
  Free games, if used wisely, can be a great promotional aid.  Things like competitions, free giveaways, all work well to increase your visibility.
• Indie Review Sites  |  Cost: $0  |  Benefit: High
  It is hard to gauge how many sales Indie Review Sites might generate, but their support is invaluable to indie developers and they do generate awareness.
• Mainstream Review Sites  |  Cost: $0  |  Benefit: Potentially High
  We didn’t really manage to crack this one.  With the exception of a Kotaku and a Destructoid news article (not a game review) we didn’t get anyone to take us on.
• Online Advertising  |  Cost: $30  |  Benefit: Low (but has potential)
  This was an inventive idea, to do a joint banner ad featuring 10 games, but the actual execution for us was probably of low benefit.  We got 6 click-throughs to the game marketplace page and an unknown number of sales.  Of course, this method does still build awareness, and people seeing the ad might have bought the game via their Xbox, or bought it a week later.  This is just the intangible nature of marketing.
• Printed Leaflets  |  Cost: $150 to $300  |  Benefit: Low
  Hmm, leaflets.  Well as mentioned, I got 5,000 printed for about $150 dollars and had some fun trying to distribute them myself before paying another $150 dollars to get someone else to distribute them door-to-door.
  In terms of return of the door-to-door leafleting, it is not likely that this was a beneficial venture.  We’d need the leaflets to generate one sale for every 11 distributed, and this is not realistic.  If 1 in 20 households has an Xbox and 1 in 5 bothers to look up the game and then 1 in 3 buys it then we’ll shift less than 20 copies.
• Seasonal Promotions  |  Cost: $0  |  Benefit: Medium
  So our seasonal promotion as described was the Valentines Day poster and press release, encouraging purchase of the game as a romantic gift.  We have anecdotal evidence that at least one person bought the game for Valentines Day.  We also got a headline on Kotaku which would never have happened with just the game’s original press release alone, and our website definitely got more traffic.  Overall we’re pretty pleased with that.
• Game Website  |  Cost: $0  |  Benefit: Low
  A game website is essential if you are to be taken seriously by the games press, and it is a good place to store any promotional material like posters and wallpapers and the like.
• Online Store  |  Cost: $0  |  Benefit: Low
  We haven’t sold any merchandise yet, so the benefit is low so far.  CafePress are expensive (even when we keep our margin low) so we aren’t expecting big sales.
• Twitter  |  Cost: $0  |  Benefit: Low (but has potential)
  If done well, I am sure this would be a fantastic tool.  If one of the twitterati were to pick up your game then that could be great.  We’re still new to Twitter so the impact was very low.
• Blog  |  Cost: $0  |  Benefit: Low
  In all honesty, this blog is not intended to sell copies of games.  It’s really just to share the experiences of game development, hopefully help other developers, and to learn some stuff in the process.
• Facebook  |  Cost: $0  |  Benefit: Unknown
  I view this a bit like Twitter, if handled well I expect it could be great, but I am pretty new to how Facebook operates and need to spend more time on it.
• Tell A Friend & Other Games  |  Cost: $0  |  Benefit: Low
  The in-game features of “tell a friend” and “other games” screen may or may not have been useful.  Without any telemetry we have no idea if anyone ever even looked at either menu option!  One thing that is clear is that trials of the game we were promoting on the “other games” screen stayed resolutely low.  Given on some days we are getting zero trials of the game we were promoting (the tragically overlooked “Horn Swaggle Islands”) it would have been easy to spot any rise .
• Radical Long Shots  |  Cost: $0  |  Benefit: None (but potentially Huge)
  Emailing celebrities I think is a great idea and I think of it a bit like buying a lottery ticket.  It’s worth one or two punts, but no more, as the promo codes are so precious.  Expect nothing back and you won’t be disappointed and the upside is potentially enormous.

So that’s it for this part, in the final part in the series we’ll try to extract the lessons learned.

For the majority of us these are alas not likely scenarios, and with little funding available it’s hard to promote your game.  This blog post is the first part in a three part series about marketing an indie game and will cover what techniques we tried.  The second part will discuss the results, so the costs and game sales and website traffic.  The third and final part will summarise what might be learned, if anything.

We’ve been doing quite a bit of practical research into marketing the Xbox indie game Gerbil Physics.  I have read what I could find on marketing on a budget and myself and the rest of the team have tried quite a few methods to promote the game; some free, some cheap but none outside a modest budget:

• Xbox Marketplace Charts – The Xbox Live dashboard charts showing “New Arrivals” and “Top Downloads” automatically feature all new games and can reach a very wide audience, albeit for a short time.  For many developers this can be the start and end of their marketing, which is a pity as there are some quality games there that deserve a wider audience.
• Press Releases – We used the august Games Press for our press releases, screenshots and promotional marketing.  As an indie developer you need to sign up and then you can post free press releases that reach a very wide games press audience.  We toyed with the idea of giving our game some weird hook, like about how many ways it could have been configured, but settled on a more conventional set of press releases.  Everything on the Games Press site is moderated and so maintains the high quality you would expect.  We looked at other press release sites like Game Release that position themselves as “indie friendly” but weren’t convinced they would be worth the 150 USD price tag.  If anyone has used their site we’d be interested to know their experiences.
• Free Games – We gave away some 50 copies of the game to various individuals and via various websites.  We ran a competition giving away free copies.
• Indie Review Sites – We submitted the game to some 40 or so websites that we’d classify as indie.  This includes sites like XBLIG and xnPlay that are devoted to Xbox Indie games and genuinely want to see games do well and get some media attention.  These sites naturally don’t get as much traffic as the mainstream sites, but they’re on the side of the indie dev and do a fantastic job of building awareness.
• Mainstream Review Sites – We submitted the game to some 20 or so websites that we’d classify as mainstream.  This includes the likes of IGN and GameSpot and a whole raft of others, including the games and entertainment columns of some newspapers.  Getting noticed here is a much tougher proposition.
• Online Advertising – This was an interesting marketing method, the brainchild of XBLIG’s Neal Emery.  The idea was 10 developers would club together and purchase a shared banner advert on some high-traffic websites.  The advert would take the potential customer to a landing page that featured all 10 games at once.  We ended up purchasing a slot for 20,000 advert appearances on sites like Official Xbox Magazine.
• Printed Leaflets – The ever inventive Pencel Games artist and designer Jon Goosens did a great job of designing promotional leaflets.  I got 5,000 of these printed and then proceeded to try to get them placed into game stores.
  The mainstream stores like Game (the UK equivalent to the US GameStop) were totally not interested and could only take promotional material if it were sanctioned by their head office.  Independent stores were more open, and some few did take the leaflets.  Interestingly, one independent store would not take the leaflets on the grounds that there was “nothing in it for them”.  I suggested that they might sell more MS Points cards but it turned out they didn’t sell MS Points cards as there wasn’t enough profit in it.
  Plan A of getting the leaflets into game stores wasn’t really successful, so Plan B was to distribute the leaflets through letterboxes in London.  I was going to distribute these all myself, but it was a lot more time consuming than I had expected and I ended up paying for someone else to distribute them door-to-door.
• Seasonal Promotions – Using the Xbox Marketplace has lots of benefits, but it is limited what you can do with your game’s price.  It’s not possible to put a game “on sale” for short periods.  As we were already at the lowest price point anyway, we had to find other ideas for seasonal promotions.  We came up with the idea of giving a game as a Valentine’s Day gift, and an amusing poster was produced to encapsulate the idea.  We then hawked this around websites and news sites asking them to publish a story about it.
• Game Website – We built a serviceable game website and there was already a vibrant art site both of which we tried to get traffic for.
• Online Store – We built a CafePress shop that allows Gerbil Physics-themed items like t-shirts, posters and buttons (badges) to be purchased.  These products really are for marketing as the mark-up is very small and we don’t really expect to be shifting boxes of stuff.  In fact shifting one that wasn’t bought by ourselves would be nice.  CafePress offer a good service though, and they deal with the production of goods, handling payments and deliveries.  All you have to do is provide the art.
• Twitter – We created a Twitter account and started building some followers.
• Blog – In all honesty, this blog is not intended to sell copies of games.  It’s really just to share the experiences of game development and hopefully help other developers.
• Facebook – We created a Facebook product page and added some media to it.  We participate in the group page “Xbox Puzzle and Trivia Games” organised by same Neal Emery who setup the online advertising group mentioned earlier.
• Tell A Friend & Other Games – We added features in-game to allow players to easily message their Xbox friends saying that they should try the game.  We also added a promo screen encouraging players to try the Indie title “Horn Swaggle Islands”.
• Radical Long Shots – One thinking-outside-the-box idea that could have been awesome, was to aim high.  Really high.  Free copies of the game were sent to TV shows.  Two free copies of the game were sent to the US’s first children.  Yes, that’s right Malia and Sasha Obama got sent some promo codes, and who knows, had they ever seen game they might have liked it.  If they have Xboxes.

Sales results and website traffic, or lack thereof, have been monitored since we launched the game on 7^th December 2009.  The next part in this blog series will give the hard facts behind sales and traffic and will try to analyse the cost/benefit of each of the above marketing methods.

The Gerbils Visit Hell

It’s been a long journey, but the “squeak-quel” to Gerbil Physics has finally made it into Peer Review today.  If all goes well we’ll be available for purchase on the Xbox 360 around the 23rd of May.  It’s been a lot of fun to make, and the gerbil protagonists certainly get about a bit more than they did in the first game.  Here is the trailer:  

Gerbil Physics Episode 2 is coming to an Xbox near you sometime in May 2010, check out some early gameplay footage here:  http://bit.ly/aVZatp

If you have 14 enemy and weapon types, with between 10 and 20 parameters each you’ve very quickly got yourself upwards of 200 floats to tweak.  In theory each float could take a very wide range of values, but their meaningful range in the game is going to be less, let’s say 10 discrete values for the purposes of this example.  Perhaps some floats will range from 1.00 to 2.00, others from 50.0 to 60.0, but let’s assume there are always 10 possible distinct values they can have in between their upper and lower bounds that are useable in the game.

200 dials to play with!

We can now imagine ourselves at a large mixing desk with 200 dials, each with 10 discrete settings (we don’t allow any to go up to eleven).  As we tune and balance the game we can imagine ourselves twiddling the various dials, a little bit here a little bit there, until we find just the right setup.  But how much twiddling would we need to do, and which dials needed twiddled?

Well, if there are 200 dials with 10 values each, that means we can have 10^200 different settings on our mixing desk.  This is a 1 with 200 zeroes on the end.  This is when the strange fact dawned on us.  “Hold on a minute, this game has more possible gameplay configurations than there are atoms in the observable universe.  How can we possibly find the configuration that’s the most fun?”

We didn’t actually know how many atoms there were in the universe, we had to look that up, but it was true!  There are something like 10^80 atoms in the observable universe.  So that is a 1 with 80 zeroes after it.  This is one of those numbers that is so outside human daily experience that is is meaninglessly large, yet it is dwarfed by our mixing desk combo possiblity!

Our little game is nothing special in terms of tweak-able parameters either.  We don’t have anything like the number of dials that many other Indie games have.  It is just a sobering and interesting fact that there are so many possiblities.

As it turned out, we couldn’t and didn’t find the “perfect” combination.  Even if there was one, how would you know when you’ve found it?  We just ended up tweaking until we found one of the many passable combinations, and we shipped that!

Here is what I believe is a fair assessment of why HSI did so poorly commercially:

What Went Wrong

• Wrong price.  At launch the game was way, way overpriced at 400 msp.  Unfortunately we got caught in the changeover from the old 800/400/200 prices to today’s 400/240/80 prices.  We went into peer review as a mid-priced game and came out of peer review as a premium priced game!  Of course, we could have pulled the game in mid-peer review in order to reprice, but we were too keen to get it released.  This was a mistake.
• Small graphics.  We didn’t properly allow for users with small HD or SD TVs.  We got lots of feedback saying the graphics were just too small and there was no camera zoom feature.  Totally our own fault for not testing on smaller screen TVs.  We added a zoomable camera in a patch, but of course that is too late for reviews and the all-important time on the new releases list.
• Overdesign.  We now know that “realistic” design does not mean fun.  We spent a lot of time coding “real” aspects of the game, where e.g. various weapon magic effects would overlap and combine in complex ways, and striving for perfect A* pathfinding at all times.  In reality these aspects are not always going to add to the fun factor, and mostly go unnoticed.  We should have taken a step back and asked ourselves were these features adding to the fun factor, or if some ”good enough” shortcut would actually be just as effective to the player and be much quicker to get working.  When a feature is complex to code, it is also complex to test.
• Long development period.  Spending almost exactly one year on development meant the market already had 4 or 5 tower defence games there, and some were not too bad.  We probably could have released 3 to 6 months earlier and would have had less competition.
• No save game.  Each level can take upwards of 20 minutes to complete.  A save game feature was something we only realised we needed towards the end of development, and by then the game was so complex it would have been very, very hard to implement.  We should have designed with this in mind from the beginning.  It would have made testing quicker too.
• Not understanding the market.  HSI is a big, hard, game, with easily 20 hours of gameplay if you replay on the tougher difficulties.  It also takes quite a while to “get into” and the 8 minute demo time I don’t think we used as well as we could have.  It is also a Tower Defence clone, albeit with some nice features.  In my opinion, most customers of Indie Games don’t want 20 hours of Xbox gameplay, even at 80 msp, as they can look to the Xbox Arcade for that.  Instead they look to Indie Games for an hour or two of real novelty that they cannot get anywhere else.

What Went Right:

• Finding a proper artist.  We employed a proper artist, and that certainly helped the overall presentation and made marketing efforts easier.  Programmer art just does not cut it.
• Plenty of polish.  There are some pretty nice visuals, animations, shader effects, transitions, that make the game feel pretty polished.
• Fun levels.  We spent ages getting level balance and presentation correct, and the end result is some pretty fun to play tower defence levels (ok, I’m possibly a little biased here, but we did get some nice feedback).
• Nice theme music.  We took an old traditional Scottish folk song and remixed it into a rather catchy, bouncy, sea-shanty type tune.
• Unlockables.  We added an unlockable “nightmare” mode if all levels were completed, and got some nice emails saying how much some people enjoyed the game.  That was nice to get, even if sales were poor.

Jesse Schell at DICE 2010

He talks about about how games are spilling out from consoles into the wider world and where this could be leading.  It is only 20 minutes long but is just packed with thought provoking facts and ideas.  Did you know Farmville has more players than there are Twitter accounts?  Or that some cars now have “grow a virtual plant” game built into their dashboard, that rewards you for careful fuel consumption by growing a flower?

Next time you take a break give it a watch.  If you are at all interested in game design its well worth your time, and he ends with a head-spinning vision of what games will be like in the not-so-distant future.

Gerbil Physics for Valentine's Day!

In an effort to market the game Gerbil Physics better, Jon the artist created a rather funny poster (high resolution downloads available here) showing the game as a kind-of Valentine’s Day gift.

The somewhat tenuous and unlikely idea was that the game would make a good Valentine’s Day gift.  This concept we hawked around some news sites, and happily got picked up by Kotaku and a few others, although sadly Microsoft declined the opportunity to feature this on their Xbox Dashboard .

I’ll be blogging about how sales were impacted by this and our other marketing efforts in a future post.