I f–king hate PPC.
There, I said it. The blasphemous words of a self-proclaimed online marketing expert. A marketing expert that is supposed to soothe you with promises of wonderful ROI and ROAS (return on ad spend) percentages and tales on how Client X was able to capture 20,000 new customers a month with a CPA (cost-per-acquisition ) of a dollar, while selling a $25 shirt made for $5.
“A dollar?” you exclaim, “That's only 4% of the sale price!” I slowly nod in unison, matching the speed of the gears behind your eyes that I already see churning. I wait for that moment, where your hopes and dreams suddenly flourish like cactus flowers in a desert monsoon, when you suddenly realize maybe this internet advertising thing isn't a gimmick after all. I smile back at your ever growing grin, as I already see the wet ink on my agency contract for the $5,000 check you'll be writing me to deliver you to the internet promised land.
Of course, that $5,000 is just the onboarding fee, because we will obviously need to do our market research, figure out the keywords and A/B test until we run out of letters in the alphabet. Once I've got you committed however, then you'll really have to buckle up for the ride, when I tell you that I'm also charging 20% on your PPC budget for my fees. “I'm saving you 50-70%, so you should be happy to spend more money!” I tell you. Now it's your turn to slowly nod along, not sure exactly what happened, but going along with the pre-printed program on the high gloss, freshly printed brochure still warm to the touch.
“Don't worry,” I say, as I clasp your shoulder, “you'll be in good hands. Now if you'll excuse me, I have some ice to go sell to an eskimo.”
Now, what I don't want you to take away from that is that online marketing or PPC is impossible or unprofitable. There's plenty of people making money on PPC out there. They just happen to be the search engines and agencies.
If I had to guess, I would say that less than 20% of PPC campaigns are profitable. Unless you are in a field with low competition, a niche market or the manufacturer, you are probably up against a giant field of adversaries with deep pockets. The likes of Amazon, Walmart, eBay and an every growing list of mega conglomerate ecommerce websites. It takes a certain level of ego (read: stupidity) to willingly go into a ring with a few good boxing gloves and square off against a Brazillian ju jitsu expert with 100 extra pounds of muscle and a 12″ longer reach than you. Now add Don King to your corner, happily squeezing 20% more margin out of you and you should get the idea of how things can go wrong in PPC.
But, I'm not here to tell you that you can't or shouldn't do PPC. I'm here to tell you that you need to realize that very smart people have trouble making a profit on PPC. It is a skill that you need to learn how to do yourself, especially if you are a small or medium sized company. Both Mike and I think of it as core business knowledge for operating an ecommerce company. If you don't know how to PPC, you might as well not know how to swim while living on an island.
So without further ado, let's take a look at some Google Adwords stats for CuttingBoard.com:
We have 5 campaigns that we have been running, the main one which is Google Shopping of course. For those that don't know, Google Shopping is a shopping comparison engine that essentially helps price shoppers find the lowest priced items. In this chart, my most important KPI (key performance indicator) is my Conversion Value / Cost. That is a ratio of how much in sales I make for every dollar spent on advertising. On Google Shopping for example, my average purchase is $114.10, the average conversion cost is $28.39, so I get $4 for every $1 spent on advertising.
The reason this KPI is so important is that it tells me immediately how well I am doing. When I factor in that my cost of goods is about 50%, then that means on paper, I am breaking even with a 2:1 ratio, as it costs me $50 in ad spend to get $100 in sales, so my gross profit is zerp ($100 sale – $50 COGS – $50 PPC = 0). In reality, I have shipping costs that are around 17% and another 3% in merchant fees, so I would actually lose $20 per sale if my ratio is 2:1.
In reality, I need to be much closer to 3:1 in order to be break even, as that figure out to be: $100 sales – $50 COGS – $33 PPC – $17 shipping – $3 merchant fee = -$3.00.
On my Google Shopping campaign, the numbers run out as: $114 sales – $57 COGS – $28 PPC – $17 shipping – $3 merchant fee = $9.00
So the reality here is that in every single one of my campaigns that isn't Google Shopping, we're actually losing money or barely breaking even! Now if I was paying an agency to run my campaign and adding another 20% to my PPC budget, I would need to be at 4:1 ratio to break even. As it is, breaking even on some of these campaigns is ok to us, because we're 1) building a customer base for follow-up orders and 2) we're building up our shipping and purchasing volume for greater discounts in the future. Unfortunately, our competitors (cough, Amazon, cough) are also thinking along the same lines and taking losses just to get people through the door.
However, this data is a little misleading as this is the average of our campaign to date. In the early stages of a PPC campaign, you'll often make mistakes on what keywords to bid on, over bidding, under bidding and other types of rookie mistakes that you hopefully catch. As time progresses, you begin optimizing better and getting the program dialed in. In our case, we actually launched CuttingBoard.com right in time for the holiday season while we were busy working on Treadmill.com and hit the ground running.
Here is a table with our Google Shopping data broken down by month:
|Date||Impressions||CTR||Clicks||Conv||Cost / Conv||Value / Conv||Conv Val / Cost||Conv Rate||Avg CPC|
A few things to notice here:
- No January stats, because we ran out of product in December and turned off advertising
- CPC rates stay mostly the same throughout
- CTR rates are fairly similar throughout, with the exception of November
- Holiday season is highly profitable in terms of ROI
- It's easy to max out Google shopping on traffic in specific niches
- May was a huge increase in performance over other months
So how do you make heads and tails out of all of this? Well, if I was a PPC Specialist, I would tell you that this sample size is still fairly small. At a few hundred clicks per month to measure conversions, variance still has a fairly large role to play. Knowing this, I'm not going to read too much into the data unless it is off by a significant margin. The volume of impressions to clicks IS a large enough sample however, so CTR should be taken as face value.
With all this said, do you care to take a guess what was the month that I popped the hood to our PPC account and began deep optimizations? If your guess was March, then you would be correct.
I noticed in February when PPC was turned back on that the Conv Val / Cost ratio had plummeted off and was in hang over mode compared to the holiday party in Nov/Dec. Conversion rates had been slammed even though I wasn't bidding any differently (as you can see from Avg CPC). My impressions were even up over December but everything suck plain sucked. The big question was why did my visitors suddenly stop buying? Most people probably sit in a circle and brainstorm all the various ways that they could be doing something wrong. If you do this, you'll find that there are a billion things that you could be doing wrong. The important thing is to actually figure out if you are doing something wrong. The solution to that problem is Google Analytics.
Here is a chart of CuttingBoard.com's Google Analytics data for ecommerce conversion, segmented by PPC traffic, organic traffic and blog traffic. Note that I segment the blog traffic from my “organic” traffic because the blog traffic is a large driver of hit and run traffic that doesn't do much for the site.
As you can see here, things get interesting when you combine this information with the PPC conversion information from Google Shopping. By looking at my organic traffic alone, I could see that I had dropped from 2.3% conversion in December to 1.03% in January and then a bleak 0.49% in February. Once again, as a data guy, I use ratios, so my December PPC/Organic conversion ratio is 6 : 2.3 (or about 3:1) and my Feb ratio is 1.8 : 0.5 (or about 3.5:1). What I'm doing is using my organic conversion rate as a benchmark for my PPC rate. If I suddenly had a sharp drop in my PPC rates but my organic conversions remained the same, alarm bells would be going off that I have a serious problem of poor quality traffic on PPC. In this case however, my ratio for Dec and Feb were fairly close, so I had to reason that shoppers just aren't that interested in spending money!
Sometimes instead of ripping out my hair and wondering why, the simplest explanation might just be the reason.
That said, you can't live on 0.5% conversion rates, so in March through April, I began a rather big overhaul of the product pages. I cleaned up the layout, added in free shipping information, put in inventory stock notices and re-arranged the layout for additional products to shop for. Conversion values ticked up on organic traffic from 0.49% to 1.29% and 1.79% to 2.86% for PPC. During this time, I also jumped from under 20K impressions on Google shopping to over 30K, even though my CPC remained the same. This was due to adding more product and also increasing the hours that I had Google Shopping running (free tip, shoppers are a waste of your time after 9pm EST).
In May, it would appear that a lot of my conversion efforts helped as my PPC conversion rates blew through the roof to over 5%, which is similar to the holiday conversion rate. My Conv Val / Cost ratio jumped to a healthy 8.5:1 and I also practically doubled my PPC conversions with even less impressions. Before totally tooting my own horn though, May happens to be the month for Mother's Day, which may simply have been a big season for cutting boards, as conversions steadily dropped after the May peak while I didn't change the website during the last two months.
I did, however, dig deep into my Google Shopping campaign and begun segmenting the products into categories and bidding on products on a granular level, so that I wasn't lumping $10 items with $100 for a blended bid. Case in point, it's pretty pointless to bid $0.80 cents for a bottle of cutting board oil when my margin is about $3 per bottle, but that is exactly the default setting should you initiate a Google Shopping campaign. I also adjusted my bids based on location (states), browser type (desktop vs mobile), hour of day, day of week and my next step will be to create bid adjustment rules.
Now, coming back full circle on why I hate PPC, the reason is that if you don't know what you are doing, every last out-of-the-box setting sets you up for failure.
- Bids are turned on during all hours of the day (3AM is the perfect time to bid… if you like cost-per-acquisitions greater than your cost of goods)
- Mobile bids are at 100% (rarely have I found anyone with mobile conversion anywhere close to desktop conversions)
- Keyword searches are on broad match. This means bidding on a broad term like “cutting board” would also net me the following traffic: cutting shears, boards of lumber, board shorts, cake cutting and other random nonsense
- Even with exact match or phrase match, if I'm bidding on “cutting board”, I will still inadvertently bid on terms such as how to make a cutting board, pictures of cutting boards, how to maintain a cutting board or the dreaded “amazon cutting board”.
- Dynamic Product Listing Ad…. if Google could invent a way to suck your PPC budget bone dry in a day, this would be it. This is like broad matching for PPC terms based on your ENTIRE webpage. Does your product say “Made in China”? How would you like to bid on the term “China Manufacturing”? Because you're about to with DPLA.
- Default campaign network is Google Search (Google properties), Search Partners (AOL, Google Maps, YouTube, Blogger) and Display Select (Adsense, third party) . YouTube is for big brands or building awareness and is hardly efficient at converting a cold lead into a hot sale. The Display Select network are third party sites that run Adsense and make money by having people click on your ads. It has and always will be ripe for abuse and will completely drain your PPC budget unless you are highly proficient.
- Location targeting: US and Canada by default. Don't ship to Canada? Well too bad, they're clicking on your ads hoping you do! The best part, if someone is China has what Google determines is the intent of searching for a product in the US, Google will be more than happy to let them click on your ad. Don't ship to China? That's too bad too.
The truth about PPC is that it is and always will be the primary source of revenue for Google. If you are the prospective gold miner , Google is the store selling pick axes. What they don't tell you is that beyond the horizon are companies with bulldozers, GPS satellites and deep earth metal detectors. Google gave you the most basic possible tool that might have worked when people could pick up gold on the ground with PPC (circa 2001-2005). Nowadays, the competition is so fierce that unless you arrived with a map to a secret location, your pick ax is woefully ill-prepared for the job.
One of the most important things you can do, is implement negative keywords. And in fact, it just so happens that we have the ultimate negative keyword list for ecommerce right here at EcomCrew. It's published for free for you to use.
Lastly, I'm going to leave this interesting article written by the Google Adwords team, about position not affecting conversions. I've always avoided the top spots because they have proven too expensive and don't provide better returns. Yet when I've spoken with PPC “experts”, I keep hearing that the #1 or #2 position receives the best conversion, in the face of the Adwords data. And lo and behold, there's always a sucker ready to bid for the top spot against all business sense… probably because their PPC agency is telling them to.
But don't worry… they're in good hands. Right?
Quick Note: We apologize for the large gap in posting and I will explain in our next blog post. There's been a flurry of activity going on here at EcomCrew and it's been keeping us too busy to post regularly. We hope to get back on track soon and keep you all updated. Thanks!