online advertising « The Direct Approach

Learn To Trust Your Instincts

Monday, May 4th, 2009

Recent personal and professsional experiences have inspired this post, which may slightly resemble a rant.

One of my dogs underwent knee surgery a couple of weeks ago to repair a torn cruciate ligament; when my husband and I picked him up, we knew right away something wasn’t right and he shouldn’t have been released from the vet’s. But it was a weekend, and we were told that everything we were concerned about could also be normal in a recovering pooch. Our guts kept telling us something was wrong, and sure enough when we finally saw the surgeon again several days later, he took one look at our dog and agreed with us. Our poor guy had to have a second surgery and ended up suffering needlessly for several days.

On the professional front, I’ve been comparing some advertising results for a client for several different sites. While I’ve previously discussed the difference between clicks and click-throughs and why you will see discrepancies, sometimes the discrepancies jump out at you enough to question them.

(more…)

04.05.2009 Web analytics, online advertising 4 Comments

Clicks vs. Click Throughs and Why You Need to Understand the Difference

Thursday, February 26th, 2009

If you run advertising campaigns, you should be receiving reports from the publisher’s ad serving software/company showing you the number of impressions your ad received, as well as the number of clicks.

This information can be valuable to you as you decide whether or not to proceed with additional ad campaigns with said vendors. But what is a click, really? For vendors and ad-serving comapnies, a click can include the following:

1. People who click through to your website
2. Spiders/bots/webcrawlers trolling the net to index information and links
3. People who clicked on your ad (by accident) and realized that wasn’t the action they wanted to take and closed things down before actually visiting your site
4. Fraudulent click activities – which end up being a combination of automated clicking devices/ip addresses who never make it to your site

As you can see there are a lot of instances where a click may not be an actual click and may be inflatign your results. This is where the click-through comes in.

A click-through is exactly what it sounds like – someone has clicked on your ad and landed on your (campaign landing) page.

It can be kind of tricky to get this information, as it requires you to have your own tracking system in place that allows you to place your own click tags in your ads (as well as those from the ad serving company/publisher.) It also requires you to have the ability to place tracking code on your site to tell when someone has clicked through from an ad.

Generally speaking, if you sit down with your IT and marketing teams, you should be able to figure out a way to do this – especially if you are using some sort of paid tracking software/system.

Did you know that the variance between clicks and click-throughs can differ as much as 30-50% in some cases? Every tracking system will be different and there will always be discrepancies between data sources, but normally you look for a less than 10% differnce.

Paid Search advertisers seem to have the lowest data discrepancies, and that is because they have long since instituted double tracking and didn’t stand for the difference between clicks and click-throughs, especially because paid search is billed on a cost-per-click basis. This forced serving companies/software systems to pay attention and fix the discrepancies quickly in order to avoid losing money.

However, in banner advertising, most programs run on a Cost per Thousand (CPM) impressions basis, so it’s advantageous to the publisher to show lots of clicks and not worry about fixing any potential discrepancies. For anyone who is running banner advertising campaigns on a CPC basis, I strongly recommend you ensure there is a second tracking option in place for you to compare the numbers.

Perhaps it’s time to stop and take a second look at your advertising campaign and see how it’s really doing.

Photo Credit: iamwahid; Stock.Xchng

26.02.2009 Banner ads, Web analytics, campaign tracking, online advertising, web marketing No Comments

Why Campaign Visitors Need Their Own Path

Tuesday, December 16th, 2008

When a prospect lands on your website from an advertising campaign, do you send them on the same path as any other visitor to your website? If you do, you need to think again.

Visitors to your site (especially) from online advertising programs have different mindsets than visitors who come to your site because they are regulars, or took the time to search for something they were interested in which brought them to your website.

Here’s what you need to know about (online ad) campaign visitors.

They were in the middle of doing something else first. It was great you caught their attention with your ad and they landed on your site, but remember this: They didn’t start out looking for your ad.

Your ad was a distraction from the original task they wanted to complete.

They want to go back to whatever it was they were doing previously, as fast as possible.

You need to understand and accept that.

You do this by making it as easy and painless as possible for them to complete the task you want them to do.

If you were to compare stats between someone who came to your site from an online ad campaign vs. someone who didn’t, chances are you would quickly come to see the following:

Campaign visitors are unlikely to go more than 1-2 pages deep on your site, whereas non-campaign visitors are more likely to explore your site and its pages.

Keep this in mind, you need to determine the single-most important action you want an ad campaign visitor to take and focus on helping that user complete that task. Giving them too many options complicates the process and makes it easy for them to bail before getting past step two, so they can get back to what they were doing before.

Anything you can do to speed up the process of completing the step you want them to take (say subscribe to your newsletter) and help them get back to their previous task, will be well received. Chances are they will come back to explore later when you’re not just a distraction.

Photo Credit: CanadaKick; Stock.xchng

16.12.2008 Web analytics, online advertising No Comments

Web Analytics – A Few Key Definitions

Wednesday, December 10th, 2008

I spend a lot of my day focused on data and analytics, and trying to determine the “why” behind actions taken in an online environment. I often forget that not everyone spends their time the same way I do, so I thought I’d take a moment to define a few key analytics measurements and describe when/how they are used or calculated.

Bounce or Exit Rate: The percentage of people leave who your website from a particular page. They may or may not have completed any actions (or visited other web pages. Bounce rate is sometimes interchanged with Single Access Rate).

Single Access Rate: The percentage of people who visited a particular page and left from that same page and did not complete any actions. This is most useful when looking at campaign landing pages or your home page.

Click Through Rate: Usually used with ad campaigns, a statistic that helps identify the number of people who clicked on a particular link, divided by the number of people who saw it. For ads, it’s usually clicks on the ad or link, divided by the number of impressions (or eyeballs). It can also be used on a page to determine the number of people who clicked on links on a particular page divided by the number of total page views of that page. Can be confused or often interchanged with conversion rate.

Conversion Rate: The number of people who completed a desired action based on the total potential audience that could take such an action. For example, if you had 100 people come to your homepage where you really wanted them to register for your newsletter, and 10 people did, your conversion rate would be 10 per cent. This metric often gets confused with click through rate or can be interchanged. In my honest opinion, a conversion rate has to do with actions, or task completion, not just clicking on a link on your campaign landing page. However, if you don’t have a newsletter, or perhaps your product is not one people can buy online, this is where many people may use clicks on a specific link also as their conversion rate.

What you might consider in that case is a specific path or journey that you want visitors to take. Ideally, you want them to visit more than one page on your site. Using a path analysis report, you can input the pages that you want people to visit and you will be able to know how many people took that specific journey. To get your conversion rate in this instance, you take the total number of people who took that journey and divide it by the number of people who viewed the page where your journey would start.

Are there any specific metrics that you’re not too sure how they are collected, what they showcase or when best to use them? Let me know and I’ll be happy to explain them as best I can.

Photo Credit: scataudo; sxc.hu

10.12.2008 Web analytics, online advertising No Comments

It’s Not Your Brand They Hate – It’s Your Actions

Monday, December 1st, 2008

A good friend and I got into a conversation a while back about actions and people. It’s not people we like or dislike, it’s their actions. It really got me thinking how that can be extended across to brands and products too. If consumers say they don’t like you – it’s often not your product or brand (unless you really do have a lousy product), but the actions taken on behalf of that product or brand by marketing and advertising campaigns, or people associated with the company.

Let’s look at this whole Motrin kerfuffle as a shining example. Motrin is the same product it always was, however, now there are swarms of mothers out there who say they will not use the product again. It has nothing to do with how good or bad the product is; it has to do with a choice that was made to run a specific advertising campaign. Because of this action, it has changed the perception people have of the brand and their feelings towards it.

If you think of the traditional marketing/advertising brief, there’s usually some sort of question around, “What do we want the consumer to think/believe after seeing our ad/product?”

And then there are questions around brand perception, asking what people currently tend to think of the brand or product.

If the folks at Motrin were to fill this out now vs. before the campaign, the answers to those questions would look a whole lot different, and if I lived under a rock, I wouldn’t know why a good many people have the perception of the brand it now seems they do.

Wouldn’t it be interesting to understand why people think the way they do? What “action” took place to put those thoughts in someone’s head?

What would happen if you broke down the actions that your company has taken over the last two to five years and started writing down the perceptions of your brand, company or product that people had before and after each action? Yes, I know there are focus groups and brand perception studies that do this now – but do they go to the next level?

What if you began tracking the perceptions against more than ad campaigns – anything you’ve done (donations, social media presence, etc.) – and began looking for patterns to understand what kinds of actions generally changed consumer’s perceptions – whether it be for good or bad… or indifferent.

You could suddenly start making a case for doing – or not doing – any number of things because of how your consumers *should* react (Note: no such thing as a sure thing).

It’s not just about asking why, it’s about understanding how and why your actions affect others.

Photo Credit: Ralph Atkinson

01.12.2008 brand advertising, online advertising 2 Comments

Why You Need to Understand Where Your Website Visitors Come From

Wednesday, October 22nd, 2008

Where do your visitors come from? I mean really come from – and I’m not referring to backgrounds or locations. I mean how do they get to your site? It definitely wasn’t magic (although we all like to think so now and again).

More likely than not some form of search (paid or organic) is going to be the largest source of traffic. If it’s not you have bigger problems that won’t be helped byt this post.

But what about after search? What is the biggest source of traffic? Is there one site that stands out on a regular basis? Or perhaps it’s a matter of “bookmarked” or “typed” urls (which is basically people opening up their browser and either manually typing in your website, or selecting it from their favourtites).

Whatever that source is – do you give it a second thought? What do you do with that information?

Do you look to see what people do once they land on your site from that particular source? Do they convert? Do they view specific content? What are their habits? Are there any patterns you can identify to help you either provide better content, help them convert or improve your overall marketing experience?

Say they come from website X – should you go spend all of your ad dollars on website X? Maybe. Maybe not. Why wouldn’t you spend all your ad dollars on that site? Well back to my earlier question of what do people do when they come to your site? Did visitors from site X convert? Did they do so at a higher or lower ration as compared to your site’s average conversion rate?

If you answered yes, then you definitely might want to consider checking out some sort of relationship with that site (if you don’t have one), after you investigate what it was that caused your traffic spike and sent people to your site in the first place.

Once you have that information – you can use it to better understand why people are coming to the site, what they are looking for and how you might work with that site owner/pubisher to create valuable content (which can also be in the form of ads) that will continue to not only drive traffic, but keep site X’s conversion rates at the same (or higher) levels.

If the conversion rate for visitors from site X was lower than your site’s average conversion rate – the question you have to ask is, is it worth it to go spend money and advertise on that site for more traffic that doesn’t convert? Perhaps. Again once you know why or how visitors are coming to your site from site X, you can then determine if you think there’s a way to improve that conversion rate.

If for whatever reason you decide you don’t think you can improve the conversion rate then perhaps you should look at focusing your ad dollars elsewhere – maybe there’s another site out there that is providing you less traffic, but a much higher conversion rate. Wouldn’t you rather spend your money focused on converting more of these individuals than by driving large numbers of unconverting traffic?

It’s not enough anymore to say “Wow site X drives lots of traffic, let’s advertise there.” Pathing tools available within any analytics system (even the free ones), if set up properly, can start to give you a better understanding of visitor behaviour and help you make more informed advertising decisions, which should increase your ROI and… and… and…

Photo Credit: afreeta; Stock Xchng

22.10.2008 Web analytics, analytics tools, online advertising, web marketing, websites 2 Comments

Online Ad Measurment: The Shorter the Better

Monday, October 20th, 2008

There really is a reason why publishers insist on animated ads stopping after 15 seconds.

Now I’m not speaking about video ads (with play and pause buttons) which apparently do work better with 30-second clips as per some study results, instead I’m talking about your traditional banner ads that are flash – or video (but look more like flash).

Many sites won’t even allow ads more than 15-seconds in length for a variety of reasons including band-width, file size and demand on their resources, but there are some that do (although I think you’ll be hard pressed to find any major sites accepting these lengths).

Working with a client on a US-based brand awareness campaign for B2B services, I was reviewing the previous fiscal year online advertising results. I found that a Fall ’08 campaign saw expandable ads with some of the creative (flash) at 15 seconds, while others were closer to 30 seconds. The only difference between the two versions were how fast the animation occurred and how smooth it was.

When it came time to refresh Spring creative, the client’s agency made sure that the ads were no more than 15 seconds and that the animation and transition of frames were very smooth so as not to run into the same types of problems that they had previously.

Comparing, results, the biggest thing that jumped out at me was that Interaction rates (the proportion of ad viewers who interact with an online ad by rolling and expanding it) jumped by 4% and brand interaction (the amount of time someone was exposed to your brand with the ad expanded) raised by a full second.

The overall theme of the campaign was the same, while it did have a visual refresh (images only), the overall concept was identical, including the wording of the ads – even how the user was asked to expand and interact with the content and how the company referred to themselves. I also checked ad placement, since in some cases accidental roll overs can be included in these rates – the ads were placed in the same sections, and used the same ratio of various ad sizes.

The only real difference was the content itself once interacted with was the main change in the ads, but a user would not know this until they expanded the ad.

I was tasked with explaining why there was such a large difference with the two ads. I had one of two theories:

1) The imagery used was either much more powerful and eye catching, which definitely may be part of the equation

2) The ads simply got to the punch line quicker. Instead of taking nearly 30 seconds to ask someone to interact with the ad, the new ads took no more than 15 seconds.

My money is on the second theory. I’m sure imagery could have had something to do with things and potentially caught someone’s eye better than the previous campaign, but what kept the attention and encouraged them to expand when the copy was the same? The fact that the user was asked to interact sooner rather than later

When you think about it, 15 seconds is a massive time difference to a user browsing web content who didn’t actually come to the site to see your ad in the first place.

Knowing that you have 10 seconds or less to keep someone on your website, you can apply the same theory in principal to your ads. We know users haven’t arrived at the website to see your ad, so it makes sense that you have a little longer than that 10 seconds to capture a user’s attention since they first have to spot your ad before you can capture their attention.

My personal recommendation? If you’re trying to encourage in-ad interaction make sure that request comes around that 12-second mark for optimal rates. That being said, it’s important to remember that testing to determine what works best for your company is always recommended.

Photo Credit: Daino_16; Stock Xchng

20.10.2008 Web analytics, online ad measurement, online advertising No Comments

Roadblocking the Way to New Online Ad Measurement

Sunday, September 28th, 2008

All forms of banner advertising should be measured with the same yardstick.

That statement couldn’t be further from the truth, yet more often than not, this is how things are done.

Roadblocking is not a new term, it’s been around in the television space for sometime now and refers to an advertiser buying all of the ad placements within a given media space. In this instance, we’re referring to an advertiser who has purchased all the impressions / ad placements on the same page of a particular website during one specific time frame.

Roadblocking is good for things such as campaign launches and general brand awareness as it gets your message out to the masses since it takes over every advertising spot your audience might see as they visit that site/web page. Click through rates for these types of ads are generally found to be low because the impression levels are so high.

Generally speaking, a roadblock has a leaderboard and a skyscraper and a box or rectangle of some kind.

To calculate click through rate for banner ads, you divide your total clicks by your total impressions. So if you get 3 million impressions as part of your roadblock campaign and 100 people click through, your click through rate through rate percentage is 100 divide by 3 million.

Which is completely incorrect.

It should be 100 divided by 1 million.

Each time an ad is shown that counts as an impression. If your ad is showing at the same time as another one of your ads in a different size on the same page, that’s also an impression for that ad – which now gives you two total impressions, for one person or set of eyeballs. And if someone were to click through on one of those two ads, that should be a 100 per cent click through rate.

Why?
Because it’s physically impossible for someone to click on both ads if served at the same time, yet standard click through rate measurements don’t take this into account.

As you evaluate your next online ad campaign, consider these two things:
1. Did you run any roadblocks?
2. Do you know if your ads ever appeared together on the same page at the same time?

If you answered yes to either of these scenarios, then you’re actually calculating your click through rate based on a physical impossibility and it’s time to see how these banner ads are really doing. You might just be surprised by what you learn.

Photo Credit: iboy_Daniel; Flickr

28.09.2008 Banner ads, Web analytics, online ad measurement, online advertising No Comments

Looking Beyond the Click Through Rate

Friday, September 5th, 2008

I’m often asked how do you know if an online ad was successful? Well many times you don’t know for sure. You can make some assumptions based on what was important to you and your organization at the time you created the ad, to measure or gauge the effectiveness it had at meeting the goals you set.

For example, if the number of people seeing your ad and coming to your site is important, than click through rates are an important metric for you. BUT in this day and age, with click through rates falling faster than Chicken Little can yell the sky is falling, that’s probably not your best metric. After all what does it really tell you? Okay, so someone came to your site. Big whoop.

Did they do anything? Did they sign up for anything? Did they accidentally click and wind up on your site? You don’t really know that do you? And there isn’t really a way to find out unless you hunt down and ask those people specifically.

So how can you tell if an ad was effetive or worked?

First – ask yourself the question of what you expect/want people to do when they see that ad online

Once you have that figured out you can start to devise a measurement strategy around that specific goal (or goals).

It might include things like signing up for a newsletter – so you would track how many people signed up (not just arrived at the newsletter page) after viewing the ad, this would likely need to be done by using a cookie or something along those lines and can/should be worked out between your ad serving company and your IT department.

If you took it one step further and had mirrored tracking links set up for each creative execution, size and even publisher (perhaps placement even), you’d be able to analyze how each version performed and determine if big boxes or leaderboards in creative A worked better and which sites got you the most registrations – or perhaps had the best ROI based on cost of media/number of subscribers.

Suddenly you can now tell a story to your stakeholders, your sales people and so on. You now know what is beginning to drive those subscriptions. Just think if you started testing the colour of the ad background, the placement of the click here etc. what you could learn about your advertising for a fraction of the cost of a print ad test.

A story is much more compelling than a click through rate. A story (usually) has a point, if not at least a beginning, a middle and an end. A CTR is just a number.

It’s also OK to have different measurement goals for different campaigns – or even multiple goals that are all measured with completely separate benchmarks. Using the same yard stick to measure everything you do doesn’t really work for much of anything – especially online advertising, so isn’t it time you shook things up and looked beyond the click through rate?

Photo Credit: miamiamia Stock.xchng

05.09.2008 Web analytics, online advertising, web marketing No Comments

Should You Care About Click Through Rates Still?

Wednesday, April 9th, 2008

The Click-through rate has always been considered an important metric in online advertising. It essentially tells you how many visitors came to your site from a particular online marketing effort (e.g email newsletter).

It used to be that it was a highly regarded metric and deemed one of, if not the most (debatable), important metrics you had to measure the effectiveness of your online campaigns.

But what does a click-through really mean? It means someone came to your site. Okay. So what? What then?

1. Did they leave?
2. Did they buy something?
3. Did they spend a lot of time on the site? If so, what pages were they most interested in?
4. Do you even know what they did?

It is no longer enough just to look at your click through rate to determine if your campaign is a success. You need to ensure that you tie all of your marketing efforts in with some kind of analytics tool (e.g Google Analytics is free) so you can understand “beyond the click.”

So a home page takeover ad drove hundreds of clicks – did that mean anything? Some say brand awareness. If everyone stayed and perused the site after clicking, perhaps, but if everyone immediately left – staying on the site for less than 10 seconds, is that really brand awareness? Maybe, but was that the objective of your ad/campaign?

I’m not saying the CTR is no longer important, it’s obviously an important part of the equation as it gives you an idea of the audience size you’re dealing with.

But from there, you need to ask yourself “Self, What did I want people to do once they got to my site?”

Use your analytics tools to find out if they did indeed do whatever it is you wanted them to do, and from there you can judge the true effectiveness of your online advertising campaigns.

It’s time to stop judging your online campaigns in a silo. Work with all your marketing teams and advertising companies to fit the pieces of the puzzle together and start to tell the story beyond the click through rate.

Photo Credit: stock.xchng/telly

09.04.2008 Web analytics, online advertising, tracking No Comments

Subscribers