Determining ROI for online advertising

In thinking this week about paid advertising, I've been thinking about my own experiences with paid advertising... not from AdWords or anything like that, but rather direct advertising on my site, and how that's worked with the companies I've been trying to work with.

When I got Busan Awesome started up and rolling, there was a point in time (actually a little later than I should have done) where I decided to contact some local businesses in Busan to advertise on my site. I started with restaurants, bars, doctors, and dentists who cater towards the expat population in Busan, based on who already had adverts on other expat websites.

The biggest problems that I had (that I did not foresee) were:

1. bars/restaurants, etc wanted to pay under the table by transfer to a Korean bank account (which I didn't have anymore), and
2. bars/restaurants, etc were stuck with the idea of "pay per click" on the advertisement, rather than impressions or other methods of determining ROI. 

While the first point posed a problem without a good answer, I'm going to discuss the second problem: determining ROI.

An ad for my website on a partner site.
How effective can it be?! If it weren't offered to
me for free, I definitely wouldn't have paid for it! 

My website is a review-based city guide website that caters to the expat that's only there for a year, or a traveler that's there for a week or so, that wants to find some cool places to go, and cool things to do. We probably already have reviews for every bar/restaurant/doctor/dentist who we would want to target, so the way I see it, for them, having an ad on our site is more of a "stamp of approval" that it's a cool, good place to hang out in. Therefore, I thought a per-impression method was a good way to determine success.

On the other hand, the bars/restaurants/etc all tended to want to track clickthroughs. They thought that if someone saw their ad and were interested, they'd click on their site. I thought that didn't make that much sense - I mean if I see an ad up for a bar on a city website, I'll maybe click on it to get the hours, but I'll tend to note that it exists, and kind of add it mentally to my "bars to check out" list, rather than click through to their website.  

So who was right? 

I think we both were right and wrong at the same time. That was a few years ago, though, and while now I'd say I know about a lot more factors of success, I still don't have a solid formula for determining ROI in a situation like this. 

There are lots of good articles out there discussing factors that you can take into account when determining ROI. On the other hand, there is no definitive recipe that exists for knowing what you're actually getting.

Take my situation that I described above for example: 

The bars/restaurants are right in that a clickthrough can definitely say "this person was interested in seeing my site." However, they're missing out on the people who get the mental affiliation between the bar/restaurant and the city guide site. Customers might say (albeit maybe not in this language): "oh yeah, they advertise on Busan Awesome. I like that site and trust their recommendations. I'll check out that bar, even though I've never clicked on the bar's website." 

So many ads! How could you not
have voluntary banner blindness?

Plenty of customers, however, could come to the site and totally ignore the ads. I know I often do. It's Banner Blindness, and there's not really a way to combat it, other than cleverly placing advertisements within an article or in some other way (such as a text link that fits into the article). 

So how DO you determine ROI from advertising?

I read a Fast Company article that talked about a study that Adobe conducted that showed that 88% of marketers said that they weren't confident that they could effectively measure ROI on an advertising campaign. That's huge! 

That same Fast Company article sites that sales conversions, value of increased traffic, and depth of
engagement are three ways that you can actually measure and determine ROI. I agree with them - those are very measurable, and that captures a big chunk of the ROI picture, but not all of the picture. 

Off the top of my head, I'm thinking about people who don't buy now but buy later, and the 'silent majority.' But still, Fast Company's list works... for companies that can measure that kind of thing.

What about the bars/restaurants that I'm trying to target? How do they measure sales conversions, other than surveying customers and asking, "how did you hear about us?" (but even then, maybe it was seeing an ad on a trusted site that brought them to the restaurant, even though it wasn't how they first heard about the website)? How do they measure increased traffic, when I already argued above that just because someone didn't click through, doesn't mean they didn't see the ad? And who actually engages with a bar/restaurant website? 

So my takeaway: 

I think the way that I need to approach advertising in the future is from a "how can I help you?" standpoint. Don't just offer an ad; offer a partnership, with posting specials, advertising some of the bar/restaurant's events, and helping out on Facebook or Twitter with some of their customer engagement. ROI still won't be able to be measured per se, but at least it will touch more people in a deeper manner.