Chasing down profitable traffic sources can be a black hole consuming all of your time and money. The problem is that almost all paid traffic campaigns start off losing money; to build a campaign, you need to spend money collecting data on what works, and then optimize towards profitability from there.But not all campaigns end up profitable either, no matter how much you optimize them. So, one of the marketer’s biggest challenges is figuring out which traffic sources are actually worth pursuing, and which ones will never become profitable and will only consume your time and money. You have a limited budget, and you need to make sure you’re making it count and not spending it on a wild goose chase. Here’s how.
How to Allocate Marketing Spend
The trick to allocating your marketing effectively is implementing some kind of traffic source triage system. When encountering a new traffic source, you want to be able to quickly and consistently categorize it into a low, medium, or high priority group, before spending any time or money testing it. You don’t have to be 100% correct for this process. After all, you never know how something is going to work until you test it. You just have to be good enough to set priorities and build up a bankroll from stable, profitable traffic sources that you can use to offset your losses from testing riskier, lower priority ones.A lot of the skills needed to accurately identify traffic sources will only come with experience. But you can get 80% of the way there and avoid the more egregious mistakes by following a few simple rules.
Three Questions You Need to Ask Yourself About Any Traffic Source
When evaluating a potential traffic source, I always ask three questions that let me easily determine how much time and money I want to apply towards mastering it.
Is there enough volume?
This question is fairly straightforward to answer. Your rep at the ad network should be able to tell you how much daily traffic you can expect from the average campaign. But you should always do your own research too.Look at the Alexa/Quantcast/Compete rank of both their homepage and their tracking URLs to get a sense how much traffic passes through the network. Do the same for any domains you see advertised on the network. Are there mostly new, low traffic advertisers(bad) or established, successful advertisers(good) on this network?If you’re evaluating an ad network, take a look at some of the publishers running with this ad network. Are this ad network’s banners the only ads on the page(good), or are you competing for attention with a cluttered field of many different ads(bad)? What’s the traffic volume of some of the publishers that seem most relevant to your product?Remember, you’re only looking to estimate volume for traffic that has at least some chance of converting for you. So if you only sell to the US, inquire about US traffic volume only.
Is the traffic cheap enough?
This question is a bit more fluid, but in general some traffic sources, keywords, and verticals are much less expensive than others.Take a look at who’s buying traffic on this network currently. What are they selling? Low margin products like commodity physical goods, or high margin financial products? How much do you think they can afford to pay for traffic? Put their domains into the Google Keyword Tool and take a look at the bids they would need to pay on search to get this kind of traffic.How competitive is this traffic source? That may mean this traffic is higher converting and more desirable, but also more expensive.
Does the traffic convert well enough?
This is the big one. If the traffic converts well enough, it really doesn’t matter how expensive it is(as long as you’re paying less per click than your CLV). Even the volume doesn’t matter as much, because you can use a very relevant, highly profitable but low volume traffic source as a laboratory to test new ideas for creatives and landing pages without fear of losing money. You can then replicate the successes to higher volume but lower margin traffic sources.You may think that it’s impossible to answer this question without actually spending money testing, but it’s actually easier than you think. You can’t answer it precisely without significant data, but you can make some pretty good guesses based on experience and past performance. Think of this a comprehensive due diligence process. To answer this question, you need to start from a broad overview and move to specific details about the competitive landscape.Have you advertised on this traffic source before? What is this ad network’s reputation on blogs and forums? Has anyone written a case study about this traffic source? Does the majority of their traffic come from wealthy countries(good) or third world countries(bad)? How do they get their traffic? Are their users incentivized to view ads(bad) or click on them(VERY bad) or do they have other quality content(good)?Then you can move to studying specifics. Who is advertising here currently, big brands(bad) or direct response advertisers(good)? Is the same advertiser buying traffic on here consistently(good) or are new advertisers constantly showing up and disappearing(bad)? Do you see the same creatives used consistently(good), or are advertisers constantly trying new creatives because they aren’t getting clicks(bad)?Check the demographics of the ad network and advertisers’ domains. Do they match the demographics of your customers? Are there any advertisers with the same business model as you(e.g. sale of physical goods, freemium, etc) running on this network? Any advertisers in the same industry? Any direct competitors?
This may seem like an unnecessarily lengthy process when the key to traffic triage is making snap decisions. But spending a couple hours studying a traffic source is well worth it compared to the months of pain chasing low quality traffic will cause you. Besides, with experience, you will come to internalize this process, and it will become second nature. That’s when you can start raking in the big bucks as a marketing consultant By the way, we’re working on software to automate answering many of the questions above. Leave your email below and you’ll be the first to get an invite to our private beta.