Playing With Google Adwords, Amused

So, I'm playing with the whole suite of google tools. Incluing site analysis, putting ads on my site, all that. And I thought, hell, why not put twenty bucks down and see how advertising works? I don't actually have some pressing need for traffic, but paying twenty buck to mess around with some heavy power tools strikes me as good times.

So here's the fun thing. You set keywords, and bids - and you see if your bid is enough to advertise "on" that keyword, as compared to others.

I put down 15 cents as my bid, and threw in some keywords that seemed to suit my site. Then I stopped, and wondered "Hey, how much are these other keywords worth?"

15 cents is enough for a click on "Dungeons & Dragons" right now. And a whole host of other stuff that's much more site-specific to me (which is boring, so I'll leave it off). It is not enough for a click on the following: "Spirit of the Century", "Minds Eye Theatre", Warhammer Fantasy Roleplay", or "Greg Stolze"

It is enough for "Castle Falkenstein" and "Amber Diceless".

Anything else I should try out?

Comments

  • edited October 2009
    That's interesting. When I Google on "Spirit of the Century" I get one sponsored link, the same one every time, for ebay. And sometimes also a second one for Century Tile.

    When I Google on "Dungeons & Dragons" I get six sponsored links, the same ones every time, just shuffled in a different order. With that much competition, I'd expect the price to be higher.

    What am I not understanding about how Adwords works?

    (I was going to suggest you try to buy "My Life with Master," because I'm curious, but it's not really a good match for your games. Maybe try some names of films? Dragonheart? Dragonslayer?)

    Paul
  • More thinking.

    Would "weyr" be a good one for you? No ads currently, but a ton of links to various Pern PBeM games. What about "Pern". Or the titles of the books?

    Paul
  • Posted By: Paul Czege
    What am I not understanding about how Adwords works?
    Adwords and sponsored links aren't the same things.

    Adwords are those side ads you see on sites sometimes that say "sponsored by google" and change depending on the words (hence the name), whereas sponsored links are just links that come up first on a google search
  • Hey Stephen,

    No kidding.

    So, a phrase like "Dungeons & Dragons," that's common across a slew of blogs showing Google ads, is cheap because Google has a lot of page inventory for displaying ads? And something less common like "Spirit of the Century" is more expensive because there's less of a universe of page inventory?

    Paul
  • That sounds like it could be a good reason. I'm not really an expert on google adwords though, so I wouldn't know for sure.
  • edited October 2009
    Posted By: Paul CzegeWhat am I not understanding about how Adwords works?
    I don't know. Whatever it is, I don't understand it either, yet.

    Edit: For further amusement, of the 19 visits adwords sent to me today, exactly 100% of them "bounced" - that is, they looked at the front page and nothing else. The "bounce rate" for the site otherwise is 51%; just over half of the people dropping in by other links read only the page they were linked to, the rest read more. Interesting, that.
  • Part of my job is working with things like Google Adwords. When choosing your Adwords try to be specific, kind of how you would if you were searching for something using Google's search engine, i.e. placing words in quotes helps a bundle. Use words that are relevant to what you think your customer wants and make sure that those words are keywords on your landing page. This means don't use Dungeons & Dragons as a search unless the page the advert is linked to says something about Dungeon's & Dragons. The cost of the advert is based on two things: how much other people are willing to bid and relevance. Get the relevance part right and you're winning half the battle. Oh and use lots of these lovely search terms at the start of your campaign.

    Another aspect of relevance is the words you choose for your advert. You're going to want to run multiple adverts. Ideally when you begin your campaign, you'll have a few adverts for each landing page (remembering to keep the searches you use relevant to the pages that each advert is linked). The wording of each advert should be relevant to its landing page and the words you used to make it show up in the first place. Which brings us back to the Dungeons & Dragons thing again. If you don't have it on the landing page or as one of the words used to make your advert appear, don't use it now. In fact, forget about using it all. The space on your advert needs to catch people's attention and call for them to click on it, but it also needs to be seen. Keeping your wording relevant will help your advert appear on page one instead of page ten. Another way to keep your advert near the top is to choose the option allowing it to only appear in the top five placements. I recommend doing both of these things.

    I said you should be running multiple adverts with an excess of search terms for one simple reason: research. After a while, you're going to have lots of information. You'll know which wording gives you the most clicks, which searches are calling up you adverts the most, and how to make your campaign more cost effective... blah blah blah. Keep collecting that information for about three months (bearing in mind that months like December might really fuck around with the results) then prune away all the shit that isn't working and I mean everything. When your three months is over, you should be running a sleek advertising campaign.

    One last thing, some people will say that paying for clicks is better than paying for impressions. I kind of agree, but that doesn't make impressions without clicks, worthless. You can't click on a television advert, radio advert, magazine advert, etc. but companies still pay large amounts of money for them. The reasons for this should be obvious. Even if someone isn't clicking on your advert, it still needs to make a lasting impression. Your impressions should make an impression!!! No need to thank me for that pun. Whether you bet on clicks or impressions, always keep your eyes on the click-through-rate. That's the thing telling you the percentage of clicks your getting for all those impressions your adverts have received.

    One more thing...

    No matter how good your campaign happens to be, familiarity creeps in and this is both good (recognition) and bad (people stop looking at your adverts). The trick is to keep it going long enough to make the thing your selling recognisable, but be willing to adapt your campaign before people stop paying attention to it. That means keeping an eye on your campaign and looking for signs that things are in decline. This doesn't mean that you delete everything and start again just because you get one bad month. You look for long term trends and when you spot them, you take action. And that means starting the whole process over again, but this time with the added advantage of having lots of delicious information to help you decide what changes to make. However, be aware that just because something didn't work last time that doesn't mean it won't work for your updated campaign. People change and so should your campaign.

    Another one more thing...

    No one can tell you what will work for your campaign. Advertising is lot of guesswork followed by a lot of research followed up by a lot of muddled interpretations of that research that is probable out of date anyway and when by some stroke of luck, people like me do get it right, we tend to be obnoxiously self-congratulatory. Conversely when we screw up, we tend to blame everyone but ourselves. Its the markets fault or the economies fault or the fault of a rival advertising campaign. Anything so long as its not our fault. The truth is that our successes are less a result of what we consciously do and our failure are more our responsibility than we would like to admit. So what does this mean to you? It means not worrying too much. It means be patient. It means not checking how your adverts are doing every day and making changes all the time.

    The last one more thing...

    Don't forget to place your adverts on websites that are relevant to what your selling as well as on Google's pages.

    The very last one more thing...

    You can use Google Adwords for text banners and image banners. I recommend starting with text banners and a small budget before throwing money and time into image banners. If you're making an image banner, I might be of some help, but I'll leave that for another discussion.

    P.S. I'm writing this while trying to reorganise my home and I apologise if I have made any spelling, grammar or punctuation errors.
  • edited October 2009
    Posted By: Levi-Who-Babbles.Posted By: Paul CzegeWhat am I not understanding about how Adwords works?
    I don't know. Whatever it is, I don't understand it either, yet.

    Edit:For further amusement, of the 19 visits adwords sent to me today, exactly 100% of them "bounced" - that is, they looked at the front page and nothing else. The "bounce rate" for the siteotherwiseis 51%; just over half of the people dropping in by other links read only the page they were linked to, the rest read more. Interesting, that.

    51% isn't a terrible bounce rate as long as you're getting enough click-throughs. Even the best sites can have high bounce rates and just because someone feels that your site isn't right for them, that doesn't mean they won't think that is isn't right for someone else that they know.

    You can help your bounce rate by making sure that what your advert says matches the content of the landing page. One reason people don't like Evony's campaign isn't because its focuses on sex, but because they feel they are being mislead. Another thing that will help is making sure you choose the right landing pages. I've seen campaigns failing simply because the wrong landing page is chosen. Your landing page should be doing the job of informing your customer about your products, giving them choices which are quick and easy to make, giving them a reason to return in the future and reasons to tell other people about your site.
  • Whoa.

    Marcus, that's some hugely good stuff. I'm kind of taken aback, given that I'm approaching this pretty casually, but I'll likely be referencing this as I do (and might be linking others here later).

    Thanks!
  • edited October 2009
    Levi, if you're not using Google Analytics, sign up for it now. You'll be able to use Analytics to find out where your visitors are coming from, which pages they are looking at on your site and for how long they stay on those pages. Using this information will help you see which pages are working and of those page which ones might be the best ones for your landing pages. It will also tell you which countries are probably the best ones to target for your campaign.

    My hunch tells me that since I write in English countries with English as a first language, e.g. United States, Australia, England, and countries with a large English speaking population, e.g. China and Denmark should be the countries my campaign targets. And this is true, but it's not the whole truth. Brazil doesn't have a large English speaking population and English isn't its first language. Despite these facts, Brazil has a strong well-developed role-playing culture which no effective campaign for rpgs should ignore. Looking at language use won't reveal countries like Brazil as viable target for your campaign. Google Analytics will reveal such countries.
  • Posted By: MJGrahamLevi, if you're not using Google Analytics, sign up for it now. You'll be able to use Analytics to find out where your visitors are coming from, which pages they are looking at on your site and for how long they stay on those pages. Using this information will help you see which pages are working and of those page which ones might be the best ones for your landing pages. It will also tell you which countries are probably the best ones to target for your campaign.
    I do have it. I am awash in information, and worry mainly that I won't actually get a damn thing written while I revel in it.
  • We were talking about marketing stuff with Paul on Thursday. We'll post some notes, soon.

    Graham
  • Posted By: GrahamWe were talking about marketing stuff with Paul on Thursday. We'll post some notes, soon.

    Graham
    Sweet.
  • Analytics is an amazing tool and a near-bottomless rabbit hole once you start digging into it. I have an unhealthy addiction to Analytics, and do some very, very deep customization whenever I set it up for a client site.

    Perhaps the single most useful thing to set up is one or more Goals -- that is, actual events on the site you care about. Typically this is stuff like "downloaded a PDF" or "submitted a form" or "bought a widget." Stuff like that. You can also set up a "funnel," which is a series of pages the users need to work through to complete the conversion. Once you have all that data, you can see at what point in the funnel folks bail out, and then go back and start patching those holes.

    Really there are books and blogs and forums and all the rest of the Analytics support community, which bears some strong resemblances to the indie gaming community. Huge amount of data. Luckily, just setting up Analytics all by its lonesome is easy to do.
  • For those that use Analytics, do you track accesses made to files that don't allow embedding of the JavaScript needed to detect their access (e.g. jpg, pdf, etc.)? If so, what mechanism to you use for this? (There are a few, mostly hacky, approaches for this, and I'm wondering if they are worth the trouble.)
  • edited October 2009
    Analytics doesn't require javascript to track anything. EDIT: Other than the JS required to run Analytics in the first place, that is. You can add code to links that sends that link data back to Analytics when the action is taken. http://google.com/support/analytics/bin/answer.py?answer=55529
Sign In or Register to comment.