The difference between "Above the Line" and "Below the Line" tracking

[Mike Sack]

"Above the Line" and "Below the Line" are terms used in the advertising business to distinguish between spends that are for branding or awareness and those that are attributable to a specific event, such as a purchase of a particular product.

Is anyone distinguishing between the two types of marketing for their online campaigns? Are there other types of marketing we should think about categorizing for the online space - perhaps a "Right on the Line" type of campaign?

The reason I am posing this question is to get a dialogue going about how to utilize tracking systems more effectively to produce better ROIs.

Share some thoughts. After this forum gets going, I'll move the discussion towards why we should pay particular attention to how we categorize spends and how best to measure them with the tools out there today.

[AussieWebmaster]

At this stage of the game I do not see branding being a real presence in search. Or more to the point I do not see companies paying to just be out there.
There seems to be some level of measurable in the PPC arena... even if it is an email address or some other non-direct monetary item.

[AussieWebmaster]

Keeping that in mind I see sales being directly quantifiable (and one type of measurable), sign ups (for whatever free things being offerred that can later be quantifiable) and the loss leader: branding.

[Mike Sack]

Most people invest in "Below the Line" spending today because you can measure it. However, if you are a major offline retailer like Sears or Circuit City, how do you go about quantifying the power of your search presence? What if key phrases for large ticket items get searched on regularly and end up driving people into the local store to make the purchase?

Those keywords might not pay for themselves, but might be worth it to your overall bottom line. I'd like to see a tracking tool that lets you track how often a particular product is viewed, tied back to the search phrases that drive those views and then correlate that data against in-store sales. I'd vary the phrases in and out of usage to see if I can make in-store numbers move at all.

Might be one way to quantify what is happening online in an offline context.

This is one simple approach. I am sure other people have some great ideas to share as well.

[AussieWebmaster]

There is the phone call tracking companies. So there are ways... landing pages, isolated domains for branding ... items only advertised online and tracked at instore sales points...

[rustybrick]

I have been pushing a famous US sofa retailer named Jennifer Convertibles to put some of its marketing budget towards search. Not necessarily to increase direct sales, heck, they don't sell sofas online. But they need to drive more traffic to their retail stores.

Its very easy to track this with analytic tools.

As I mentioned in this thread, you can set up goals and track to see which keywords are converting on those goals. So if a goal is using the "store locator" then you can track that. In addition, many companies have 800 numbers just for the Web calls, so if they call, you know the lead was delivered by a Web visit.

[AussieWebmaster]

I seem to recall Volkswagon using set colors for the 'bug' that were only advertsied online and on TV (two colors)... another good way to isolate measurement offline or by media.

[Mike Sack]

Here's a thought -

With Google about to release some powerful tools for distributing your message to targeted geographical locations, does anyone think this will help establish some linkages between online and offline sales?

Suppose we ran a test where we distributed (re: Rustybrick's example) listings for sofas to a geo area served by a single store.

Then we measured the impact of search and looked at store foot traffic. Could we make a correlation between search traffic and in-store traffic patterns? Could we then apply this to the larger context of advertising across all of the supported geographies?

My take is we'll be able to use local search to do this. One geography may not be enough - we may have to test several in order to see impact, and, of course, larger geographies are likely to produce more quantifiable results as well.

Once we understand the impact of search to offline traffic, we can start to figure out how much each click is worth...thereby justifying a possible increase in spend. Let's face it - the real money will flow into search when we can connect the dots for marketers...online to offline.

Anyone else have some thoughts on this?

[K.S. Katz]

I believe that "Above the Line" is extremely important for my company. We do online press releases as a major way to get our name out to the community. With PRWeb.com, you can target specific geographic areas much like Google AdWords.

For "Below the Line" we have a complicated backend that tracks phone and web leads. Each campaign has their own 1-800 number and landing page.

Interesting enough, our largest lead generator is word-of-mouth from existing clients.

[doppelganger]

Originally Posted by AussieWebmaster

At this stage of the game I do not see branding being a real presence in search. Or more to the point I do not see companies paying to just be out there.
There seems to be some level of measurable in the PPC arena... even if it is an email address or some other non-direct monetary item.

Actually, most people running PPC search engine ads are not tracking beyond the click, so they are just driving traffic, not an action. According the info earlier this year from the SES NYC show one of the Nielsen folks gave the following numbers for 2003 PPC campaigns:

55% = drive traffic
24% = awareness
19% = drive sales
1% = positioning

Keep in mind as well, there were some great numbers that came out this year about brand recall in search... search beat out other online ad formats, such as banner ads, by a 3 to 1 margin. That means it's a good channel for branding.

[AussieWebmaster]

Originally Posted by doppelganger

Actually, most people running PPC search engine ads are not tracking beyond the click, so they are just driving traffic, not an action. According the info earlier this year from the SES NYC show one of the Nielsen folks gave the following numbers for 2003 PPC campaigns:

55% = drive traffic
24% = awareness
19% = drive sales
1% = positioning

Keep in mind as well, there were some great numbers that came out this year about brand recall in search... search beat out other online ad formats, such as banner ads, by a 3 to 1 margin. That means it's a good channel for branding.

I think the tests were skewed... It is like the ones that say you get more clicks from the number 1 organic listings when I know for a fact that is not the case across a few industries...
That top Adword grabs them every time.. or at least 2-3 times more than the free search spot does.

[Mike Sack]

Originally Posted by AussieWebmaster

I think the tests were skewed... It is like the ones that say you get more clicks from the number 1 organic listings when I know for a fact that is not the case across a few industries...
That top Adword grabs them every time.. or at least 2-3 times more than the free search spot does.

Do you have an research that support this claim? The evidence I see (from our customer base) suggests that paid search does get as much or more "clicks" from the engines, but that organic listing produce far more conversions - almost a 5:1 ratio. Even when we track deferred conversions over time, we still see a 3:1 ratio of conversion for organic vs. paid.

[Rob]

I think any SM campaign is miguided if all that is measured is the conversion.

Brand building at the early research stages is important as well, so while a broad term may not convert, it does get your brand out there and recognized so that when the customer gets close to the decision making stage, if we've done our jobs right as search marketers, our clients site will be found at that stage as well.

As part of a focus groups study my company did we found one case in particular (although there were many other similar cases) where a customer wasn't even aware of a particular brand when they started their research, but at the end they were determined that that brand was in fact what they were looking for in the first place (she was looking for a specific type of running shoes).

So, while it would be difficult to impossible to measure this as a successful conversion, it was a conversion none the less.

[AussieWebmaster]

Originally Posted by Rob

I think any SM campaign is miguided if all that is measured is the conversion.

Brand building at the early research stages is important as well, so while a broad term may not convert, it does get your brand out there and recognized so that when the customer gets close to the decision making stage, if we've done our jobs right as search marketers, our clients site will be found at that stage as well.

As part of a focus groups study my company did we found one case in particular (although there were many other similar cases) where a customer wasn't even aware of a particular brand when they started their research, but at the end they were determined that that brand was in fact what they were looking for in the first place (she was looking for a specific type of running shoes).

So, while it would be difficult to impossible to measure this as a successful conversion, it was a conversion none the less.

But was this really branding or well presented information to help the decision process. Careful optimization and good content (a conversion tool).

[Rob]

Originally Posted by AussieWebmaster

But was this really branding or well presented information to help the decision process. Careful optimization and good content (a conversion tool).

Good Point Aussie, it was primarily good placement, however it took a few searches before she really decided to check into the brand. It was only after it showed up a few times, in a few different results, from different sources before she decided to check it out.

Therefore it may not necessarily have been the brand building they did, but it was still a successful conversion, just not one easily measured because in this case she went to the store to buy them. She wanted to touch them and try them on before committing to the purchase.

I guess that was my point in the beginning, that using online conversions as the only measure of a campaign's success is limited.

[Jeff Nienaber]

No matter where the sale occurs, online or off, branding is real. At a minimum, a consumer must view a brand 7 times before any trust is established and a relationship is forged. SEM provides a cost effective platform for valuable branding impressions to potential customers regardless of a businesses' ecommerce capabilities. Not to mention the brand osmosis that occurs when a company is served in a serp, or essentially qualified, by the users favorite search engine.

[Mike Sack]

Originally Posted by Jeff Nienaber

No matter where the sale occurs, online or off, branding is real. At a minimum, a consumer must view a brand 7 times before any trust is established and a relationship is forged. SEM provides a cost effective platform for valuable branding impressions to potential customers regardless of a businesses' ecommerce capabilities. Not to mention the brand osmosis that occurs when a company is served in a serp, or essentially qualified, by the users favorite search engine.

Jeff makes a really interesting point - that a customer must view a brand 7 times before any trust is established. Does increased placement or visibility of your search campaigns impact conversion? If this theory is correct, getting multiple listings to appear in the same result set for a given query becomes extremely important.

Some of my own research has examined the likelihood of click-through and conversion when a customer has a Google Adword listing, an Organic listing or both. Anyone care to guess what produces the greatest likelihood of conversion?

[rustybrick]

Originally Posted by Mike Sack

Some of my own research has examined the likelihood of click-through and conversion when a customer has a Google Adword listing, an Organic listing or both. Anyone care to guess what produces the greatest likelihood of conversion?

I would venture a guess that the clickthrough rate if you have an organic and paid listing on the same page is about 90%.

The conversion really depends on a lot of things. Its interesting. I recently started tracking for some of my e-commerce clients, the time it takes for a customer to make a purchase.

Here is a snapshot of some data in time:

Time to transaction placed in number of days:
0 days 48.65%
1 days 1.80%
2 days 3.60%
3 days 3.60%
4 days 1.80%
5 days 0.00%
6 days 0.00%
7 days 0.00%
8-14 days 5.41%
15-30 days 27.03%
31-60 days 8.11%

Time to transaction placed in number of sessions:
0 days 48.65%
1 days 1.80%
2 days 3.60%
3 days 3.60%
4 days 1.80%
5 days 0.00%
6 days 0.00%
7 days 0.00%
8-14 days 5.41%
15-30 days 27.03%
31-60 days 8.11%

Data is very new, so take it with a grain of salt. Plus I really can't say which e-commerce site this is from.

[Jeff Nienaber]

Jeff makes a really interesting point - that a customer must view a brand 7 times before any trust is established. Does increased placement or visibility of your search campaigns impact conversion? If this theory is correct, getting multiple listings to appear in the same result set for a given query becomes extremely important.

Thanks Mike. Your question is important and in my experience the answer is a resounding yes. I've tested several PPC ads alongside of high-ranking organic listings and watched both benefit from the dual placement. This speaks to the heart of complete SEM: strong SEO, strong PPC, and strong PI (if it's right for you).

[AussieWebmaster]

Originally Posted by rustybrick

I would venture a guess that the clickthrough rate if you have an organic and paid listing on the same page is about 90%.

The conversion really depends on a lot of things. Its interesting. I recently started tracking for some of my e-commerce clients, the time it takes for a customer to make a purchase.

Here is a snapshot of some data in time:

Time to transaction placed in number of days:
0 days 48.65%
1 days 1.80%
2 days 3.60%
3 days 3.60%
4 days 1.80%
5 days 0.00%
6 days 0.00%
7 days 0.00%
8-14 days 5.41%
15-30 days 27.03%
31-60 days 8.11%

Time to transaction placed in number of sessions:
0 days 48.65%
1 days 1.80%
2 days 3.60%
3 days 3.60%
4 days 1.80%
5 days 0.00%
6 days 0.00%
7 days 0.00%
8-14 days 5.41%
15-30 days 27.03%
31-60 days 8.11%

Data is very new, so take it with a grain of salt. Plus I really can't say which e-commerce site this is from.

90% is a little high.. I have instances of 2-3 organic listings for 2 companies and the 1 and 2 spots in PPC and I don't get a 90% CTR!!!

And don't say I need to write better copy etc... LOL