While ad testing, proper account setup, and conversion tracking are necessary for a successful account; you can’t get any data without a competitive bid.
If your bid is too low; then you ad never shows. If your bids are too high, then you can quickly lose a lot of money; and if you are new to PPC, then you might decide that PPC doesn’t work and abandon the medium.
With mature accounts with lots of data; bidding isn’t too difficult. It can be as simple as last click attribution, or as complex as using multi-touch attribution models. However, the reason you have options is because you have the data.
A new account doesn’t have any data. How can you effectively set bids based upon what you rely on the most – the data – if you don’t have any?
In this column, you’ll look at a few bid methods that can help get you started setting your initial bids.
Setting Bids Based Upon Estimated Conversion Rates
This model attempts to simulate traditional bid models by guessing first at your conversion rates and then using your typical bid formulas to set bids based upon your goals, such as CPA or ROAS.
Estimating conversion rates can be very difficult. If you offer a free white paper, your conversion rates can vary from 1%-20% based upon your form design. Many companies like to use 2% for ecommerce. However, its common to see ecommerce rates for less expensive items hit 5-10% and for more expensive items be around 0.1%-0.5%.
If you have historical analytics data, you that to help you make some good guesses as to your PPC conversions. If you have worked in a vertical before, that can help as you’ll. However, all you are trying to do is reach a reasonable assumption of a conversion rate so you can use your regular formulas to set the initial bids.
So, do some educated guesses and research to come to an estimated conversion rate, and then use that rate to set the bids based upon your CPA or ROAS targets. The method is fairly straightforward.
The advantage of this method is that you usually don’t lose too much money (and hopefully make some) when the account first launches. The disadvantage is that sometimes you launch an account and the traffic is so low that you immediately have to redo all the bids. That brings us to the second bid method, ensuring that you get clicks for the brand new account.
Accumulate Lots of Clicks to See What Works
In this model, you’re looking for the most clicks possible so you can determine what is going to work and what will not work. By having some data, you can focus your future energy working in the correct places.
So, your goal is lots of data, clicks, and traffic regardless of how it converts so you can make decisions moving forward.
This does not mean you still should not spend time creating you’ll organized ad groups – you must give the system good inputs to get good outputs and useful information.
The end goal is data first and profit a distant second for the initial bids; later on those goals can change
In this model, you can use the estimates from the AdWords Keyword Tool or from the Bing Ads Intelligence plugin to get some starting bid information and use that in the accounts as the starting bids.
What will happen as soon as you put those bids live on AdWords is that some of the keywords will automatically be ‘below first page bids’; even though that was the bid suggestion in the tool. You can then use bulk edit function to raise all the keyword bids to the first page bid.
A note of caution, whenever using a bulk edit tool to raise all keywords to first page bids – always put in a bid cap (upper limit or CPC bid limit in Google’s interface depending on the screen you are working with) so you don’t have a few keywords at $100+ bids.
If you really just want lots of clicks; you can even go a step further and use the budget optimizer for the campaign bid setting:
I find that budget optimizer (AdWords will set my bids to help maximize my clicks…) setting is great for publishers or for just collecting lots of clicks regardless of the quality of the click. However, as its goal is to get the most clicks possible, if you have a few broad keywords (especially if they are broad matched), those few words could get all your clicks so that your data is not evenly spread throughout the new campaigns.
In the end, with this bid method – you’re taking suggested bids from existing data sources and using them as your starting bids; so there is no guess work involved.
The advantage of this method is that you will get lots of data. The disadvantage is that you might not make any money at all; so be prepared to lose some money while you determine what the bids should be based upon your target goals.
The Hybrid Approach: Traffic & Conversions
In this method, you want to estimate conversion rates so that you are hoping to make some money from the initial launch; but you also want to use the existing data from AdWords & Bing (using their cost estimations) to ensure that you will get some traffic.
Start by estimating conversion rates and determining your starting bids. Next, you’ll need to see if those bids are close to page one bids. If they are, then you can go ahead and set the bids live. If they aren’t then you need to reassess your goals.
For instance, if you do the math and guess that you can start the bids at $1 based upon your estimated conversion rates; but the AdWords tool says almost all the clicks in the industry start at $5-$10; then you might have a problem to overcome. There are some simple questions you need to ask yourself:
- Can you realistically raise the conversion rate?
- Can you raise the target CPA?
- Can you raise the average order value?
- Do you have a business model problem?
Personally, I always go through this bid exercise even if I’m not going to use this bid method as it will point out business model issues. I’ve seen times where the target CPA isn’t possible in an industry and you need to reevaluate how you are making money, the value of a lead, your upsell opportunities, and lifetime customer values in order to be able to be competitive. It’s best to evaluate the business model and how the advertising needs to be executed before spending money.
The advantage of this method is that it will give you a realistic expectation of traffic levels and conversions. The downside is that it might scare you away from keywords that could work you’ll and it can be limiting if you don’t take a few chances with some keywords (which could be at a later date) to see if high bids and competitive keywords do work for the business.
The other downside is that this can be a lot more work prelaunch; but I’ve never thought this exercise was a waste of time. It’s good to have an overall idea of conversions and CPAs before you start.
Just Bid High, Build History, and See What Happens
Another starting bid method is to bid into high positions, ensure you build up a great history for Quality Scores and for the account, and then make adjustments after you get some really good data.
This bid method is generally used by three types of people:
- Novices who are new to AdWords and heard bad advice
- People who engage in ‘ego bidding’; and tend to lose a lot of money doing this
- Skilled experts who are bidding this way for a very specific reason and have set aside a budget of money they will lose in order to establish the brand and ads in the search results
The only group that should be using this method is number 3, the skilled experts. The other two should change their starting bid method.
This method is not for small budgets nor the faint of heart. There have been accounts we purposefully set aside $25,000-$100,000 as ”play the game money” just to establish a strong history and presence. The money can often be recouped down the road through the sheer volume of clicks that occur at the top of the page.
If you have a small budget, don’t try it. If you are new to AdWords or PPC in general – don’t try it.
The reason it can be useful gets into some interesting quality score theory. There are some of us that believe that Google does not normalize CTR for quality score purposes very well when examining the top versus the side of the page. This group does believe that Google normalizes CTR for quality score by position; but that there are some issues.
For instance, among the ads in the top of the page it is normalized well among other ads in the top of the page. Among the ads on the side of the page, the CTR is normalized well for all the ads that are on the side of the page. However, when comparing top versus side ads, the normalization often gives preference to top of page ads as it underestimates CTRs that occur on the top of the page, and thus giving some of these keywords higher quality scores than they deserve. Its through top of page exposure that you end up with entire accounts that have 8-12% CTRs through ad testing and keeping most of the keywords (assuming they are profitable) at the top of the page.
Of course, this is only a theory; and one that cannot be proven or disproven without understanding Google’s actual formulas. However, it seems to be true from an outsider’s perspective.
The advantage of this bid method is that you can build up a great history so that over time, it will cost you less to maintain these top positions that can bring in a significant amount of traffic. Before you ever try this, you must have very well segmented account and great ads – an exceptional account build is necessary to try this bid technique.
The disadvantage of this bid method is pretty obvious,it can cost a lot of money to bid this way and it will not always work. It is a gamble. However, for those skilled in the game of PPC and have the budgets to work with – it is a good way to establish a presence.
For many companies, playing a long tail or geographic strategy is a better idea than a high initial bid strategy. I add the information here in hopes that more people will realize it is not always necessary to overbid and to caution anyone before attempting to try this on their own. The majority accounts do not need to start this way; but I commonly hear the advice so I wanted to address it and put it in perspective of other initial bid types.
If you have comments on this bid method, or another columnist who wants to write an article disproving this bid method – I welcome the input and debate (I don’t claim to be right – I just claim it’s a theory that has worked for me for many accounts).
Know Where You Want To Go
One of the tricks to bidding and creating the accounts is knowing how you want your PPC account to mature. If one of your goals is to get to a point where you can try CPA bidding, then you’ll first focus on getting as many conversions as possible, establishing a conversion history, and then letting Google do the work. This means that you will start with a highly targeted narrow keyword set and try to get traffic and conversions immediately after account launch.
If you are launching a new product, and your initial focus is to drive product awareness, then often you want lots of clicks and impressions across a wide variety of keywords, and then in a few months you’ll switch to a more ROAS focused approach. This account would use a lot more modified broad match and broader keywords to attain those goals than the account focused on conversions.
In the famous words of Lewis Carroll: “If you don’t know where you’re going, any road will get you there.“
You need to know where you’re going, and setting initial bids is just the beginning of the journey. However,set the initial bids based upon the road that you do want to travel.
There are many ways to set initial bids; including ones I didn’t even mention such as the $1/$5/$10 method in which every ad group starts with a semi-competitive number just to see what happens or the ‘instinct method’ where you’ve been doing PPC for so long, you’re just good at guessing some initial bids and don’t even bother to look up any numbers before setting the first bid.
Personally, I feel the hybrid approach is the best method for most companies. It will help to ensure you can get some traffic and to see if you can hit your target CPAs with your keywords, bids, ads, and landing pages.
When you first start accounts, you usually make few adjustments in the first week to month (depending on account size) as you are accumulating data to determine bids, negative keywords, tests, etc that you need to set bids that have a chance of giving you the data necessary to work with so that later on in the account’s history you can make data-driven decisions. Picking one of these solutions should help you find a way to set initial bids so you can take your PPC account from the beginning of the journey to a place where you are making informed, data-driven decisions.