Pay-Per-Click Won’t Save You

(Originally published on February 8, 2013)

I’ve never run a big business, but I have spent (correction, wasted) tens of thousands of dollars in my startups on Pay Per Click (PPC) ads through Google Adwords and other ad networks. I’ve gotten some business from it but I’ve never been able to make it work. Maybe there are some companies out there who are wildly successful with Adwords, but I’ve never met anyone who has built a business on PPC and seldom have I even met anyone who has built a profitable portion of their business on PPC. And I’m certainly not going to use it with my new startup, MoonClerk.

I’ve tried PPC in a variety of ways:

• Learning it and running the campaign myself
• Hiring a digital agency to run the campaign
• Hiring a freelancer/consultant to run the campaign
• Hiring Trada, a firm that specializes in crowdsourcing campaigns to experts

PPC Results For My Businesses:

Zekten

The Idea:

Zekten Autos (now defunct) was a startup of mine that failed for a lot of reasons. If you’ve ever heard of the used car chain, CarMax, you know they buy your trade-in even if you don’t buy a car from them. I was going to do this but do it solely online. People would submit details on their cars, I’d give them a rough quote of what I’d pay, send a driver to inspect and pick it up, and then sell it wholesale. With the PPC campaign, I was trying to get people to fill out information on their car to receive a quote.

The Approach:

With Zekten Autos, I learned Google Adwords and implemented a PPC campaign on my own. I read a ton of books, researched online, watched instructional videos, went to a multi-day training course, and got some personal one-on-one training from a PPC expert. I ran the campaign for a few months.

The Problems:

1) Starting and then managing a PPC campaign can be extremely time-consuming. For a one man operation, it was overwhelming to try to optimize a campaign. I didn’t have the time or experience to optimize the campaign. There are so many details and ways you can make your campaign better and so many changes to the platform, that for a PPC campaign to be really successful, you almost need to have someone focused on it.

2) My campaign was regional and didn’t scale. I could only send drivers out in a radius of a couple hundred miles or else it didn’t make financial sense. When I reduced the campaign to this radius, there simply wasn’t enough volume to bring in the business that I needed, even if I had higher conversion rates.

3) My business was early stage so I didn’t know my conversion funnel. I didn’t have enough data to optimize it. In hindsight, the form that people had to fill out to give information on their car and upload photos was extremely long. But I was in the early stages and it took a while to figure this out. The clicks I was getting weren’t that expensive but my conversion rate was so low it didn’t matter. I also really didn’t know if the business had wings in and of itself. Even if I had run the world’s best PPC campaign, it would be hard to tell because I hadn’t validated the business — there were a lot of variables at play. Unless PPC was immediately and randomly some sort of gangbuster success, it wasn’t going to validate the business for me.

I ended up spending thousands of dollars in advertising dollars and training (probably between $5,000 and $10,000.) Almost all of that was completely wasted from a monetary standpoint. Still, it was a great learning experience because by studying Adwords myself, I was able to have intelligent conversations with other firms about it down the road.

Rice Bowls

The Idea:

Rice Bowls is a nonprofit. We weren’t trying to get people to straight-up donate. We ship out little plastic piggy banks in the shape of a bowl of rice. We give them away to leaders of groups (teachers, pastors, etc.), they hand them out, people fill them up with change, then they send the funds back to us and we fund the food budgets for our 52 orphanage partners. With the PPC campaign, we were trying to get people to order the free bowls.

The Approach:

We hired a digital agency to run this campaign for us. They didn’t specialize in PPC but they had a digital marketing team with experience in PPC. We were using them for some other stuff (non-advertising related) so we decided to use them for this advertising campaign. We ran the campaign for a few months.

The Problems:

1) We had a pretty good idea of our conversion funnel — people order the bowls online and we know it works. We’ve been around since 1980 so we have a good idea of how many people actually use the bowls they receive, how much money gets filled in each bowl, how long it takes them to send the funds in, etc. So, we knew what numbers we needed to hit. We just couldn’t ever hit them. It always cost us way more to acquire a donor (person who orders bowls) than the projected revenue from the bowls they ordered.

2) People don’t generally do web searches to donate or figure out how to help nonprofits. Or, at least, they don’t follow through with their intentions. Generally, they have a connection to a nonprofit or a particular cause. They seek out specific nonprofits from personal referrals. We ended up getting a lot of low quality clicks that didn’t actually convert. The clicks weren’t that expensive but they just didn’t convert.

3) The way we structured the manner in which we payed the digital agency didn’t really provide any incentive for them to optimize our results. We basically paid them cost +10%. We set a budget and monitored performance and they adjusted as we went along. I won’t make that mistake again. We ended up spending thousands of dollars and never actually came out ahead.

Bellstrike

The Idea:

Bellstrike is a for-profit business that lets nonprofits set up good-looking, donation-enabled websites in about 1–2 minutes. It operates on a hybrid freemium business model. WePay, the third party payment processor that we use, charges 3.5% on any donations. We charge an additional 6% but cap our fees at $80/month. With the PPC campaign, we were trying to get nonprofits to sign up for our service.

The Approach:

I did a lot of research on ODesk and ELance looking for a PPC consultant. I narrowed them down by experience, reviews, performance, etc. I emailed some questions, and then set up Skype interviews with the top candidates before finally choosing one. From a customer service standpoint, working with the guy I chose was a great experience. He was super helpful, quick, and knowledgeable. I paid him a set fee to set everything up and then a set monthly fee to optimize. His fees were very reasonable (in the hundreds of dollars to set up and under $100 to optimize every month.) I ran the campaign for about 10 months.

The Problems:

Even though my geographic limitations were pretty large — the entire United States — there just weren’t that many searches for the inexpensive keywords that we needed to bid on. There are some large, expensive players in the nonprofit software space and they’re able to pay up the nose for expensive keywords. Because Bellstrike is a hybrid freemium model, we can’t pay those prices. Combining our revenue model with the fact that we were so early stage, we didn’t know what our Lifetime Customer Value (LCV) was. That made it difficult to even figure out what we needed to pay per conversion. Over the course of the 10 months or so that I ran the campaign, we were able to get the cost per conversion down but weren’t really able to scale to where we were picking up more than a handful of customers a week. Come to find out, the LCV of my individual nonprofit customers isn’t that high, so these additional customers didn’t really help from a revenue or profitability standpoint. We still haven’t “broken even” on those customers and I’m not sure we ever will.

Overall, I spent about $4,000 — $5,000 on Google Adwords for Bellstrike. It did bring in some revenue but, so far, nowhere near enough to cover the costs. My best customer acquisition method so far has been plain old word of mouth.

Loft Resumes

The Idea:

Loft Resumes provides well-designed creative resumes for non-designers. Job seekers pick a design, upload their content, and then we have our graphic artists custom-typeset the content into the design and email them a PDF. With the PPC campaign, we were trying to get people to order a resume.

The Approach:

I did some PPC research on this one and was super pumped about the opportunities. There were a ton of resume/CV keywords, almost all of them had low competition so they were inexpensive, and they had a ton of search traffic. Additionally, we could go global with our campaign (though we ended up not doing so) because we have a global market. I also did a lot of research about how I would approach this market with PPC. I wanted to make sure that I gave it a great shot, that I put enough dollars into it to really be able to test it out, and that I gave it enough time. I ended up going with a firm called Trada that specializes in PPC and has a unique model for optimizing campaigns. Trada assigns paid search experts to your account based on their past success with campaigns similar to yours. You set a budget (they required a minimum of $3,333 every month for 3 months,) and set conversion goals (eg. “we are willing to pay $30 per new customer.) Their experts build and optimize your campaign and make a commission on the difference between your conversion cost goals and what they were actually able to get a conversion for. In my experience, the process was completely transparent and granular — I could always see what was happening, all keyword, all ad groups, clicks, conversions, etc. If the discrepancy between your conversion goals and what they are able to convert for grows, they adjust the conversion goals downward so you aren’t overpaying in commissions. We ran the campaign for 3 months.

The Problems:

1) With Loft Resumes, we know our Lifetime Customer Value. Our average purchase is about $115 and that’s also our LCV. Having this knowledge and knowing our exact profit margins let us calculate exactly what we were willing and able to pay for a conversion. However, we don’t really have the potential for repeat customers. Pretty much everyone purchases 1 resume design and that’s it. Most e-commerce companies can reasonably assume that they may have repeat customers and can potentially amortize customer acquisition costs over time. We have to do it immediately, which is harder.

2) We were able to get essentially unlimited clicks. So, scale wasn’t a problem. And, those clicks were pretty cheap — there wasn’t much competition. The problem was that our conversion rates from PPC were abysmal. We knew that our other traffic sources had decent conversion rates and we’re built on Shopify so we follow their checkout process — it seems to work for a lot of their customers. So, the problem wasn’t usability. The low conversion rates were likely due to the fact that people weren’t actively searching in order to purchase resumes. They were searching for resume ideas, or resume inspiration, or whatever else. But they weren’t thinking when they searched, “I need to pay for a service that will design my resume.” Our service was pretty new and unique (the only one that we knew of at the time) so it wasn’t really possible to attract “post-intent” customers, meaning we didn’t have the possibility of attracting customers who were already thinking of purchasing what we offered.

We spent a little under $10,000 for this experiment. On the orders that we did get, our average customer acquisition cost was around $200 — $300. Not great when your average revenue per customer is $115.

MoonClerk

The Idea:

MoonClerk is a new startup of mine that came out of beta and launched to the public on Monday of this week. MoonClerk lets anyone immediately start accepting recurring payments with a branded checkout experience and without any code required. It’s geared mostly toward small businesses such as of-the month clubs, service providers (eg. lawn care services.), nonprofits, daycares, gyms, etc.

The Approach: Don’t do Pay-Per-Click advertising.

The Problems: See the “Estimating Your PPC Potential” section below to find out how I discovered that PPC wouldn’t be a good marketing/sales channel for us.

What I’ve Learned

Who PPC Can’t Work For

• Low Price Freemium Businesses (e.g. Bellstrike)
• Businesses that need “pre-intent” marketing (e.g. Loft Resumes)
• Early stagerecurring revenue businesses that don’t yet know their lifetime customer value
• Early stage businesses that haven’t optimized their conversion funnels
• Local/regional businesses that want large scale (e.g. local spas probably won’t have enough search volume for their target keywords to make or break their business)
• Businesses that don’t have the time or resources to manage their PPC campaign
• Businesses that have low search traffic target keywords and that produce low volume sales (e.g. Bellstrike)
• Businesses that have high search traffic target keywords with a lot of competition

Who PPC Might Work For (these businesses may need a combo of the following factors to be successful)

• Businesses that know their lifetime customer value and whose value extends beyond one purchase (e.g. high monthly volume B2B SaaS company with a long customer lifetime)
• Regional businesses with high dollar volume sales and keywords that can obtain high conversion rates (e.g. local DUI attorneys).
• Businesses that have already optimized their conversion funnels
• Businesses that have the time to manage, study and optimize performance or have the money to pay someone to do all of this for them
• Businesses that have products that people are already searching for with the intent to buy (“post-intent” search)

Estimating Your PPC Potential

One good thing that’s come out of all of this is I’ve now got a quick back-of-the-napkin way to estimate if there is any potential for PPC in any of my new ventures. Having this knowledge let’s me know that I don’t even need to consider PPC for MoonClerk. Here’s how it works:

First, I follow the rules listed in the blog post Your Market is Smaller Than You Think by Rob Walling. Rob’s “Bottum-Up” 4 Steps to finding the true search volume for certain keywords helps me get a good estimate on search traffic. Then, I do a little math. In this example, I’ll use a Loft Resumes target search term — [Creative Resumes] and related keywords.

10,000 This is my estimate of the local Monthly Searches for [Creative Resumes] keywords (using Rob’s methods)

x .5 (true number of searches my ads will show up for as per Rob’s methods)

= 5,000 Potential Monthly Searches

5,000 Searches

x 2% Click Through Rate (this is an estimated global average CTR across Google Adwords — it could actually be unattainable for any particular situation, but we’ll go with it)

= 100 Monthly Clicks

100 Monthly Clicks

x 2% Conversion Rate (This is a rough average across the web for e-commerce and other businesses. It’s very rough, but we’ll go with it.)

= 2 Potential Conversions per Month

100 Clicks

x $2 per Click (this is the average cost of a click as per Google Adwords estimator with regards to [Creative Resumes] keywords. I’m usually not conservative here because you can probably get a price 25% lower than what is listed)

= $200 Monthly Cost

$200 Monthly Cost

/ 2 Conversions

= $100 Per Conversion

So, I can guesstimate that it will cost me roughly $100 to pick up a new customer. For Loft Resumes, my average customer brings in $115. With my overhead and costs added in, I’d lose money going down the PPC route. In reality, with Loft Resumes, almost none of the search traffic was coming from “post-intent” purchasers. That means the conversion rates were much worse than 2% for these type keywords. I ended up paying $200 — $300 per conversion and lost money.

That, my friends, doesn’t feel good.

Originally published at blog.doddcaldwell.com.