Advertising is brutally simple, at least in concept:
- Come up with a message
- Blast that message on a particular ad medium
- Measure the results
- Refine and repeat
The trouble is, most advertisers struggle with steps 1, 3, and 4. Coming up with a message requires research and testing. Measuring results requires analysis. Refinement requires all of the above. As a result, most website advertising isn’t cost effective.
However, to be fair, most advertising in general isn’t cost effective. Here’s what you need to know if you’re going to make advertising work.
Separate Ads By Purpose
There are basically two types of ads (by purpose):
- “Branding” ads, which don’t try to get the consumer to take action. Instead, they present and promote your company brand.
- “Direct Response” ads, which try to encourage the consumer to buy, to visit the website, etc.
Advertising for branding purposes has it’s place. But this article is focused on direct response advertising. Digital advertising lends itself very nicely to direct response, and with a good process any advertiser can achieve quality results.
Five Rules For Every Advertisement You Run
There’s nothing like a simple set of rules to help ensure your advertisements are successful. While there’s some simplification in our rule set, all the basics are here.
Rule #1 – Create an Ad For Them, Not For You
The biggest mistake most advertisers make starts with the message. A traditional sale flyer is a good example:
A sale flyer like this serves the business goals, not the consumer’s goals. The message with a flyer like this is “we have good deals.” While consumers love a good deal, they’re concerned about other issues too (part selection, installation advice and support, easy returns, etc. etc.). Consumers also need to see how a part (or part retailer) is going to help them with their bigger problems. Many consumers are worried about vehicle reliability, for example. An ad that addresses reliability concerns – perhaps by saying “We have great prices on parts from reliable, well-known brands” – might perform better.
To make a belaboured point: Research your customers. Learn their concerns and create ad messages that reflect these concerns. For example:
We know that consumers who buy parts online have two pain points:
- They don’t like to pay for shipping
- They don’t like to wait days for a part to arrive
These ads are addressing consumer concerns – not trying to help move an overstock of WD-40®.
Rule #2 – There’s An Unlimited Supply Of Advertising Opportunities, Most Of Which Don’t Work
Auto part etailers can buy AdWords or Bing PPC ads, Facebook ads, Twitter ads, Pinterest ads, Instagram ads, forum sponsorships, all sorts of publisher sponsorships, traditional magazine ads, club memberships, radio ads (online and offline), etc.
Most of these ad opportunities won’t work. Some won’t work because they’re overpriced. Others because they don’t reach your target consumers. Others because you just can’t find a message that resonates.
The point: You have to test your advertising options, and you can’t be surprised when they fail. Therefore:
- Never enter into any long-term ad buy without testing it first
- Before you try an ad placement, get all the info you need to predict performance
- Don’t bother with ads that can’t possibly be cost effective (more on this below)
- Don’t try ads unless you know how the platform works
One of the most expensive mistakes any etailer can make is to buy ads that don’t work. Spending a pile of money on ads that have no impact is the same as lighting said pile of money on fire. Don’t do it.
Rule #3 – Your Ad Performance Is Only As Good As Your Testing Process
Which of your current ad placements generate a positive ROI? How is that ROI changing over time – is it getting better as you refine your ads, or getting worse? What percentage of your test ad messages succeed?
Obviously, in order to answer these questions you need:
- A good system for measuring results.
- A process for testing new ad copy/creative.
If you’re selling parts online, the good news is that measurement (#1) is reasonably straightforward. Call tracking and Google Analytics can take you pretty far.
As far as testing ads (#2), etailers should focus on ad platforms that allow for testing. This means that simple ad buys – like buying banners or forum sponsorships – are sub-optimal. Publishers who offer simple ad buys don’t have a system that advertisers can use to refine their ads. That means testing and improving these ads is very hard.
AdWords, Facebook, Bing, etc. all offer testing tools advertisers can use to improve their ads. That makes them a better choice for advertisers who don’t have their own in-house testing tools.
Rule #4 – You Can’t Know Until You Test
Ad salespeople, business owners, experienced staff, and even marketing consultants all believe that they can spot the difference between a good advertising opportunity and a bad one. While there are some things you can do to predict ad performance, no one knows how an ad placement is going to work. The only way to know is to test.
Basically, this means that advertisers should keep an open mind. Successful advertising starts with experimentation.
Rule #5 – Always Buy Results, Never Buy Impressions
Here’s a trick question – which ad buy is a better deal?
- $1,000 that yields 10 million impressions and 10 sales, or
- $1,000 that yields 10,000 impressions and 100 sales
Unless we’re really interested in branding or exposure (and we aren’t), option #2 is the way to fly. 100 sales is better than 10. However, the companies selling ad space usually want to focus you on impressions. This is wrong.
Ad impressions aren’t what matters – it’s actions that you can trace back to your ads which are important. Therefore, when it’s time to buy advertising, you should project results before you write a check. Assume a best case and worst case response rate, do the math, and see how it backs out.
Example: You can buy a month’s worth of forum sponsorship for $100. The estimate is that your ads will get 200k impressions. However, these are banner ads. Response rates on banner ads are going to be somewhere between 0.01% and 0.02% (or at least that’s a good estimate).
The best case scenario is that you get 40 clicks on your banners (0.02% click thru rate * 200,000). The worst case is that you get 20 clicks (0.01% * 200,000). Therefore, you can project about 30 clicks as a result of your sponsorship. Would you pay $100 to get about 30 clicks?
Projecting possible best case and worst case outcomes before you buy any advertising is a best practice. Advertisers who don’t sit down and do the math usually pay way too much for advertising.
While there are a lot of little things to know about advertising, the big things are pretty straightforward: Know your audience. Measure everything. Test everything. Buy smart. Do these things, and you’ll be ahead of a lot of the competition.
Another great piece of advice – one that is perhaps a bit self-serving: Hire an ad management company to help you. For the cost of a small ad buy, you can bring in a team that knows the industry up and down and that can help you with measurement, analysis, etc. Often times, the cost of bringing in ad professionals is lower than the expense of wasted ad dollars.