My blog has moved!

You should be automatically redirected in 6 seconds. If not, visit
and update your bookmarks.

My name's Joel Kelly and I live in Halifax, NS.

I'm a 20something guy doing digital and social media strategy for a Halifax-based marketing agency.

I'm a vegan nerd and marketing asshole.

You should follow me on Twitter.

Contact me about whatever (like, say, your marketing questions) at

Thursday, May 29, 2008

All engine-less cars are equally useless

No matter how pretty the design of your Flash-only, single-page website is, it's just as useless at a 1994-era static, ugly website.

If you don't allow direct, internal linking, or image-copying, or text-quoting, your site is useless. If you build a website, assume someone is going to want to share it with their friends, and likely only a single part. If you don't think someone will want to do that, then why are you making it in the first place? Make it as simple as possible for someone to spread what you have to say and offer around. If you don't, if you force people to start at your splash page and proceed from there, you deserve no one's time or indulgence.

That website is no better than an ancient relic of the internet, a page with frames and manually-updated HTML. A useless site is a useless site, no matter how pretty.

A Ferrari without an engine is just as useless as a Chevette without an engine.

The only difference is that you could probably sell your useless Ferrari for more money.

A bad static website is just as useless as a dynamic, pretty, single-page Flash site. But you can probably charge your client more money for the latter.

Wednesday, May 28, 2008

Sell by the CPM

Seriously, don't try to get me to buy ad space on your website based on time. I'm just going to assume you're trying to trick me into paying more than the space is worth.

Internet advertising is measurable and trackable, so I'm going to buy based on measurable and trackable amounts. I'm going to buy based on how many impressions the ad actually receives. So if you insist on selling by the month or week, I'm still going to work out the effective CPM (cost per thousand impressions) and decide from there if it's a good deal. And I'm never going to buy space on a site that can't give me real data on how many impressions the ad is likely to receive. So save us both that trouble and sell on a CPM basis from the get-go.

Now, I understand that it might be a little more complicated for you. Because how do you decide how valuable those thousand impressions are?

Well, first consider supply and demand. If you have huge inventory, you don't want to risk it going to waste and not having enough advertisers buying, or enough impressions being purchased. So you'll want to set a CPM that gets most or all of your impressions purchased.

If you don't get a lot of traffic, you'd better have a great reason why those limited eyeballs are really valuable. Define a niche for your site that makes that small number of impressions worth buying. If you can justify it, you might even be able to set a reasonably high CPM. If you're selling out your inventory, then you're doing okay (if you're selling out your inventory really, really quickly, maybe you should up your CPM).

If you're in the mid-range, with lots of impressions but not more than you know what to do with, you don't necessarily need a mid-range CPM. Again, if you can justify why those impressions are more valuable than the next guy's (if you have a niche website, or serve a need that no one else is covering), then you don't have to charge a middle of the road price.

So if you're selling ad space on your website, those are a few things to keep in mind. And remember, the best thing you can do to get yourself on one of my buys is this: Make it painless. If I can call you up or fire you off an email and expect great service and great ROI, and provide a great experience for myself and my client, then you've got the best shot at getting on a buy.

At the end of the day it's about value for my client, and since they're getting billed based on how much time I spend on a buy, too, if buying on your website gets me valuable impressions without having to spend a lot of time on it, you're doing us all a favor.

Monday, May 26, 2008

Reach ain't everything - Tip #1

My last post on web marketing threw out an example of something that I sometimes need to explain to clients when preparing online campaigns for them: Reach vs. % Comp UV.

Reach ain't everything. Just because a certain site might hit a huge percentage of your audience (Yahoo!'s network of properties, for instance, reaches about 100% of internet users in Canada), that doesn't mean those sites are sufficient on your buy, or even the best way to spend your money.

Sites with huge reach often have prices to match. While Hotmail may have more impressions than they know what to do with and can therefore offer a low CPM (cost per thousand impressions), most sites with huge reach charge a big price for their valuable impressions.

So what do you do with a small budget, or a client extremely concerned with efficiencies? That's where % Comp UV comes in. It's the percentage of a site's unique visitors that are composed of your target demo.

So one site, say, Facebook, might hit 90% of your target demo, but their % Comp UV may only be 3%. However, a smaller, more targeted site might only reach about 5% of your demo, but it could be composed almost entirely of people in your demo. Find enough of those smaller sites with huge % Comp UV and you might be able to spend less over a bunch of sites and hit all the same people without having to dump all your money on one or two sites with huge reach.

Of course, if you're breaking your buy up into smaller chunks like that you'll want to make sure you're not hitting the exact same people over and over again on those sites.

Agencies use tools to measure reach, % Comp UV, and duplication, among many other metrics, but a good eye and intuition (and carefully worded questions to the people running the sites you're interested in) can still get you a long way. I can't really talk about specific tools I use at my day job, or go into detail about how we construct buys, but hopefully the little bit of information I've given will create a bit of awareness of what goes into buying media online.

Sunday, May 25, 2008

What, would you say, you do here?

It's odd that I haven't really discussed what I actually do in web marketing for a living.

I think it's because I'm not entirely set on where this blog is going, yet. It's been reasonably professional and painfully formal so far, which bothers me a bit. I'm no stranger to overly-personal blogging, so I'm hoping I'll be able to find a comfortable middle ground soon.

Anyway, my day-to-day job is planning and buying online advertising campaigns. So I get briefed on the client/product/creative and I plan the online execution from there. I find the sites to purchase impressions on, make the buys, monitor campaigns, and work with the client to keep them up to speed. So that includes a lot of research, negotiation, and more spreadsheets and acronyms than could be easily imagined. Oh, and meetings after meetings after meetings...

That's for the display side. I also handle all the search marketing for our client's campaigns, both PPC and SEO. So I would set up and monitor and maintain Google/Yahoo paid link campaigns, and make SEO recommendations.

The rest of my job is being the internet expert in the office. Lots of meetings explaining social media, why CTR isn't as important as people would like to think, explaining the difference between reach and %Comp UV, and why if you make a website with a loading page you should be fired from the internet.

I'm actually preparing an hour-long talk that I'll be presenting to all my coworkers on Thursday about Web 2.0 and Social Media. I'm going to try to have it recorded, so depending on company policy I might be able to post it online.

So, yeah, that's a little bit of what I do. And also I'm the one male in the media department, so the other part of my job is being "the guy" who tries to turn the conversation away from Girlicious and toward Indiana Jones. With limited success.

Friday, May 23, 2008

Ask, Don't Tell

Everybody hates meetings. Well, most people do.

If you love meetings you're likely organizing lots of them, so you should probably keep in mind that most people hate meetings.

And the people who like meetings are typically the ones doing all the talking, the ones doing all the telling. The people who hate meetings are typically silent, thinking about other things, likely thinking about all the work that's piling up while they're away from it.

If you're stuck in a meeting, the best way to get something out of it, or at least try to keep it relevant, is to ask good, tough questions.

How many times have you sat in a meeting and wondered why the conversation has taken this turn? How often do you think no one's actually addressing relevant issues? When this happens, ask the question!

If you're in a meeting about how to promote a product before the product has even been developed, before there's even been a creative brief, ask why. Ask how you can be planning an advertising campaign before you've even differentiated the product from the competition. This sort of thing happens, and if nobody asks the question everyone will be wondering long after why nobody asked it way back then.

Ask the question. Best case you'll get a good answer and everyone will be enlightened. Worst case you might not get dragged into so many meetings anymore...

Monday, May 19, 2008

Out of the Loop

It's remarkable, the anxiety caused by being out of the loop, which itself was caused by anxiety.

The past few weeks have been so stressful and depressing that I haven't had the energy to even watch tech news, let alone keep up with all the blogs I need to be reading. And I do mean need. My job sort of requires that I keep up-to-date with SEO/M news, and my position in the company is defined by my ability to keep on top of these things, to be the guy who knows what's going on online.

And right now I have no idea. It's paralyzing, I'd say, to stare at all these RSS feeds without any idea where to start. Knowing that there isn't enough time for me to get caught up, knowing that I'll just have to start fresh, and hope that nothing really important happened last week.

If anyone knows any big news that someone involved in internet marketing and social media should be aware of, I'd really appreciate you dropping some links.

Monday, May 5, 2008

Your Social Networking DNA

So I'm at the AIM Conference and I just heard a presentation by the General Manager of Yahoo! Canada talk about Social Networking DNA.

His point was, basically, it's unreasonable to expect social networks to continue to operate as they do now, because currently they're the only space where people of all ages interact together. He gave the example that if he started shopping for clothes where his daughter shops, she'd throw away her clothes.

Sure, that makes sense.

And I agree that it's weird and probably unsustainable for Facebook to draw people of every demographic.

And I agree that the future is going hyper local. Everyone will have their own networks where they can interact with the people they want to interact with, when they want to. There will still be Facebook, where everyone is, but there will be smaller, more niche networks, where you can talk to specific groups of people with whom you have a specific connection.

But I do not agree that people will have a single profile, a single "you" that will get carried from network to network, saving you from creating new profiles and logins everywhere you go.

Why? Because you're not the same "you" on every network. Your Facebook "you" is different than your Twitter "you" because your Twitter messages are broadcast to everyone. You're just not the same person everywhere you go.

And if someone steals your one, single social networking identity, that could be as damaging to your business and reputation as your real, actual identity.

Having a single "you" is as bad as being forced to use biometric information to gain access to physical locations or computer data. If someone steals your credit card and your identity, you cancel the card and you contact the police and the government, because you'll be able to prove through all your other "profiles" that you're the real you.

If someone steals your fingerprint, or gets the data from your iris and manages to steal your identity that way, how could you possibly prove you're the real you?

If someone steals your one, single social media DNA, you're in the same boat.