Article #17.5-Ad Data Analysis

I strongly recommend reading KSAC #17-Advertising Tips & Stats first if you haven't yet.

Welcome to part 2 where we get to the nitty gritty for the information and stat junkies.

Here I'm going to openly display our full advertising stats and together we'll take a look and draw some conclusions.

Fortunately Calliope Games was cool enough to be open about these stats too so you could benefit from their experience as well. So let's start with The Titan Series' BGG ad data, courtesy of Ray Wehrs and Cassidy Werner of Calliope Games.

Calliope Stats 1

I want to start with this one, because even though it's the biggest ad campaign we'll discuss (over 5 million impressions!), it's really concise and there is a lot we can pull from it in an easy to digest manner.
(Note: BGG gives you "Bonus free impressions" on the homepage if you spend enough. That's what the right section is, for the same dates established on the left. You'll see that a lot in the following charts.)

Taking a look at the yellow rows where the campaign was in Preview stage. (They advertised their preview page like we did for ADAPT 1.0, even more so actually, which is why, of course, that we reached out to them).

  • Notice the Click Through Rates (CTR) in the far right columns. That percentage starts high, and drops regularly and substantially by the time the campaign was launched.
  • Look at the total number of clicks that went to a Preview Page where a pledge could NOT be made. Compare that total to the green live clicks; it's less than 1/3rd.
  • Observe the power of the Bonus Homepage Impressions. The CTR is WAY higher than that of the standard run. Note: These clicks will draw CTR away from the other run, but the value is clearly greater.
  • Look at the bottom section, and the purple highlighted # of clicks. Yes, it's smaller than most other "Clicks" in this image, but these clicks were all generated in ONE DAY, not One Month. The BGG Homepage "Takeover" (a BGG special buy) was stated by Calliope to be "The single best piece of advertising their entire campaign", and it was only 1 day. Perhaps we should all consider it...?

Calliope was kind enough to share their stats advertising on TGN and through AdRoll too. Here they are.

Calliope Stats 2

These happen to be combined into large Preview Page and Live Campaign sections instead of by month.

  • For TGN, you can see the CTR drops tremendously again from Preview to Live.
  • Yet, for AdRoll the CTR climbs! That's because AdRoll is a "Retargeting" platform. It is their job to make your ad more effective over time, not less (by use of keyword and platform retargeting). Perhaps they might make advertising a preview page worth it... though bear in mind Calliope uses this net data to suggest otherwise.

That's a great view of a Seasonal and Monthly analysis.

Now let's look at DAILY data.

Starting in the past, and working our way forward, first I present our Data on The King's Armory ad campaign. It's a great jumping off point because it was a solid 30 day campaign for a successful Kickstarter, features stellar CTRs, and even used mostly animated flash ads.

I've split the chart wherever we updated/refreshed our ads, and therefore temperature-highlights are focused within each ad-block. Yellow dates are "Coming soon" ads.

The King's Armroy Stats

Some takeaways:

  • The ad campaign was hugely successful with .40%+ CTRs the whole coming soon week, and .30%+ the 1st and most of the 3nd weeks of the campaign. This is a credit to the ads: their content, their presentation, etc.
  • Updating your ad updates can be delayed (notice our Launch Day Ad posted 2 days late).
  • Data can get lost, though rare (10/29); OR may well not have been live at all on this day.
  • Updating an ad must mean it's down for a large portion of the day (or meant this in 2013) as Impressions plummeted around when we updated everytime. (Good to know, right!?)
  • Glitches can keep your Homepage Bonus from showing up (see coming soon week above).
  • We reduced saturation (lowered Impressions) mid campaign to focus fire on the naturally exciting times. This is why the 11/7 through 11/14 week has far lower impressions by day. This seems to have worked for us.
  • Number of Clicks is clearly reasonably proportionate to the number of Impressions. More Impressions = more clicks. 11/15/2013 is a stand-out example of this.
  • Level of CTR is inversely proportionate to the number of impressions (as the same people are seeing the ad they already clicked on). 11/18/2013, the Monday after we funded when we made a big push all in one day, and 11/27/2013, our last day, are stand-out examples of this. Low CTR, but huge Clicks. (Ultimately the same point as the above bullet.)
  • Pounding more advertising WILL result in more visibility for your campaign.

Here is our Halfsies Dice Campaign Daily Ad data. You'll see a short formal preview time, which I consider fairly successful since it was only 4.25 days in advance, and then the daily data for the entire 34 day campaign.

Halfsies Dice Stats

Some Takeaways:

  • Overall these CTRs were much lower than TKAs. There are a number of reasons this could be and they're all conjecture...
    • The campaign was clearly for dice. It's not like a game that you might be interested in or might not be. You know you want some new rollers or you don't. (Note that Halfsies had 2.5x the number of backers TKA did, so this is a pretty telling bit of conjecture.)
    • The adverts were animated like TKA, but they were not as interesting as TKA's banners. Mostly just new dice showing up, replacing the old, to display multiple over time without overcrowding the image.
    • The season was different. If this matters in anyway, I don't know, but people do spend much more money in November each year than in March.
    • It was 1.5 years later, and the market flooded between TKA and Halfsies. I wonder if CTRs across the board lowered on BGG in these years...?
  • We again see a huge CTR jump every time an ad was updated.
  • Uht oh! Where are our Home Page Bonuses? - I never got them! So check your ad stats mid-campaign to make sure all is going to plan.
  • Biggest of all: Look at the purple fields. What power the "48 Hours left, 8 Dice sets unlocked" banner has! And the "Final Day-All Stretch Goals Unlocked!" banner has the power to create the single highest CTR of the campaign.

Here is our ADAPT (1.0) Data, but there's a lot less to take from it as it was so very short. I actually pulled all ads 5 days before formally pulling the Kickstarter Campaign to save money when I knew a cancelling was inevitable. The campaign didn't never ran long enough to update the ad.

ADAPT 1.0 Short run Stats

Ads pulled 11/2, campaign cancelled 11/7

Possible Takeaways:

  • Your ads may not go up as early on the day you requested as you hope (Day 1 has terrible impression count as a result.) This can have lasting damage on the bandwagon/snowball effect that Kickstarters thrive on, so be sure send in your ads several days early and get confirmation they received them.
  • CTRs looking better than Halfsies could reinforce my every point above regarding Halfsies CTR being lower than TKA's (timing, schedule, season, etc).
  • Those homepage bonuses are amazing CTRs, and since the impressions were more consistent it provides slightly better data to pull info from.
    • Note that when added to the regular run's CTRs it creates a fairly balanced tally.
    • Halloween has another amazing day for BGG CTRs. I honestly wouldn't have expected this in the first place, let alone to show some level of consistency.
    • The art we could fit in the square Homepage bonus ad was killer, and that surely helped.
  • The impressions count started huge because we planned on manipulating the quantities daily, while maintaining an $800 plan; but then plummeted when we pulled 85% of our advertising cutting our losses when we knew were looking at an imminent cancel.
  • A singular freaking jump in CTR on a Wednesday is encouraging, but that

The Biggest Question: Are certain days of the week better?

Well looking at...

Click to enlarge

Click to enlarge.

and
click to enlarge

Click to enlarge.

...we see a pretty large jump on TKA's Monday's average clicks, but that's largely due to 11/18/15 (red) throwing the data off and without that it shows as the WORST day like the Halfsies data; otherwise we see a noteworthy curve toward Thurs-Sunday for TKA, but that's largely due to the huge 10/31-11/3 push. Halfsies is pretty balanced save for Sunday.

I don't believe this data lends evidence that one day or another can be counted on for more clicks reliably, but it does begin to indicate that Thurs-Sunday is pretty decent and Monday just might be the worst. (Makes one wonder then why Mondays seem good for pledges, while the weekends tend to slump.) But remember, this is only "clicks".

Now looking, rather, at CTR, the more reliable stat, we see a pretty balanced week for TKA aside from a notable dip on Wednesday; while Halfsies shows a Thursday and Sunday as peaks.-Looking at this data I'd conclude that every day of the week is reliable for getting clicks, with Monday & Wednesday indicating a lull, and Thursday indicating a boon.-Whether those clicks result in pledges or not is a whole different story. For that I'd study Kicktraq's "daily data" which shows consistent jumps on Mondays, and lulls on Sundays. If anything, that would be what answers the question "what days are best to advertise".

Here's just a quick snapshot of what I mean there.

Kicktraq Daily TKAThis data here shows that for TKA Monday (Cyan) and SATURDAY (Dark Green) were the best days for both backers and pledge income. The last day is Neon Green, and the last 3 hours (Red) kicktraq treated as a new day due to timezone issues; I merge that data with the last day when thinking on it.

Kicktraq Daily Halfsies

This data here again favors Monday (Cyan), but prefers FRIDAY (Yellow) over Saturday; again the last 3 hours in red ...that's A LOT of pledges in the last 3 hours!

Data By Country!

Finally, I want to show some very interesting "Data by Country". The takeaways here bear far less on how you advertise, or even who you advertise to (as I don't think you can control it on BGG), but it shares a lot of international data about where your biggest markets are.

I put the Takeaways right on each campaign's column, for ease of association.

Net "By Country Data".

Brace yourself, this image is huge...

Gate Keeper Games by Country TKA & Halfsies...no really, click it to view the whole thing and/or download it.

Possible Takeaways:

Impressions:

  • USA, UK, Germany, and Canada are your biggest markets.
  • Followed by Poland, Spain, Italy, and Australia.
  • Of 196 Countries reporting, 125 of them aren't likely to click your ad (approximately 0.9% of your impressions).

CTR:

  • Afghanistan apparently really wants board games.
  • China has the absolute worst CTR imaginable (esp. given impressions), my fulfillment data also says nobody there orders anything either.
  • Several countries (Columbia, Peru, Slovenia, +) burn away a ton of your impressions.

Misc:

  • Vanuatu is a place. You learn something every day.
  • Poland is a fun place to live.
  • There are places that you DO NOT want to advertise in. So let's look at...

Top Countries to Advertise in (Geotargeting)

...based on CTRs and actual order rates.

Taking into account three campaigns, the numbers are very consistent.

Countries worth advertising inNotes:

  • This is the condensed list from 196 countries BGG has reached for us down to the top 34.
  • If a country isn't listed it's because the CTR was terrible; the CTR was poor given the number of impressions it was chewing up; or the CTR was fine or even outstanding, but across the successful campaigns there were consistently zero-to-one orders.
  • The countries highlighted in Green will give you the most traffic and the most business.
  • The countries listed in Red are the ones I'll be kicking off the list in my future Geotargeting based on a lack of order-to-click ratio. Look at Brazil & Taiwan's impression counts ... compare to zero orders. What could those 25,000 impressions do in the USA, UK, and Germany?
  • All this Geotargeting said: I think it's wise to avoid excessive Geotargeting at the start and close of your campaigns. This way if there's someone truly interested, they'll get a chance to see it without the country eating up the impressions.
  • This isn't personal, it's business.

PHEW!

Wrap Up

Amidsts all this data, there's surely something I'm missing. If you find anything of note in these tables and charts, please share it with us below. And if you found this article helpful, please use the links in the right sidebar to share it with someone you think will appreciate it.

John Wrot!