An Introduction & Best Practices To DoubleClick Bid Strategies & Automations

Chris Hodgson, 17th June 2016

Depending on the size and portfolio of your client, bid management software with automated bidding tools can really be worth their investment, saving time and money.

Any accomplished search specialist will tell you how only weeks, even months of manual testing can provide you with the right insight to develop a robust bid strategy.

Many will test product categories individually, making incremental weekly changes to build a clear picture of how those small groups of keywords perform in different positions and at higher/lower bid thresholds. This is still an effective method of optimising your account for optimal performance, but can be agonisingly slow.

As one of the fastest moving industries, search marketing demands fast results, thus increasing the need for more automated strategies and less time spent running tedious reports that result in the same conclusions.

Businesses that are looking to save time spent gathering these essential learnings will undeniably benefit from a bid management solution platform equipped with automated bidding technologies.

Advertisers using DoubleClick’s latest release (v3) can take advantage of the custom bid strategies tool which allows users to specify what actions or metrics they want to grow across specific AdWords accounts synced with the platform.

Bid Strategies Best Practices

Step 1 – Choose your bid strategy

The source of many advertisers’ struggles when it comes to selecting a bid strategy is more often than not the ability to identify what is going to deliver the highest value and how this fits with your overall objectives. Getting this wrong could result in a cascade of poor results, higher cost per conversions and lower Return on Investment (ROI).

For example, if your client measures your performance on ROI and not on volume of sales, then this advertiser is more suited to a position based strategy which focuses on brand protection and helping to maintain maximal coverage for low cost, high converting or high AOV terms.

Conversely, if your client marks your performance on volume of sales and traffic rather than ROI or Cost of Sale (CoS), then this advertiser is more suited to a bid strategy which looks to maximise frequency of ad delivery for keywords which have a history of driving sales at a given position or bid.

Combination bid strategies aim to balance volume and efficiency by setting a max cost per conversion/action you are happy to pay with the overall goal of driving as many sales as possible within the stated CPA.

How does DoubleClick know how to do this?

DoubleClick’s sophisticated learning algorithm aims to identify similar attributes in users entering a live auction to past converting users, increasing exposure for users who are considered more valuable and more likely to complete the desired action.

In a nutshell, the real value of implementing bid strategies is the time saved in reacting to daily, hourly and even seasonal trends. For instance, if Monday is a peak day for your client and usually on this day bids are raised to achieve a certain level of coverage then DoubleClick bid strategies will use historical data to decide if and when these changes are appropriate.

Step 2 – Configure your bid strategy

The DoubleClick interface has an easy step by step setup for bid strategies allowing you to specify:

  • Accounts, campaigns, ad groups, keywords and product groups participating in the bid strategy
  • Goals you want to achieve, this can be:
  • Conversion
  • CPA
  • Clicks
  • Revenue
  • Position
  • Combination
  • Constraints such as bid floor, ceiling, and position

The tool can also look to achieve goals using custom formulas and goals using custom attribution models.

Once the above is implemented, the algorithm will begin observing past and current user attributes to accurately target higher value prospects. Bids will be raised for those considered most likely to convert but no higher than your stipulated max bid.

If the tool does not have enough data, or cannot achieve your goal within the bid range stipulated, then a warning will be given in the ‘bid strategy health’ column. It is then the advertiser’s decision to either increase the bid or relax the cost per conversion threshold until enough data is accumulated.

Bid Strategy Attributes

Bid strategy attributes can help advertisers identify the reason why their keywords or product groups within the bid strategies folder aren’t performing optimally or within range of their stipulated cost per action.

Percentage without clicks / impressions / conversions

Bidding strategies reporting on a high ‘% without clicks’ is a strong indicator that your max bid is too low or your cost per action is unrealistic, so use this to re-evaluate your thresholds or remove keywords which you’re not happy to pay above your ceiling bid and replace with new keywords to test.

Percentage at max bid

This refers to the percentage of days within a given timeframe that keywords within the bidding group are reaching their effective maximum bid set.

If you consider your cost per action for this bidding group to be palatable, then the correct course of action to take in this example is to increase your maximum bid for added volume (budget permitting).

As with the percentage without clicks example above, if your cost per action is already reaching its stated limits then you may want to consider removing keywords with high CPCs and reintroduce new keywords for further testing.

Percentage of optimal cost

Expressed as a percentage, the proportion of the actual cost in line with your stated cost per action with 100% being right on the mark.

So if you’re achieving low % values for this attribute then this indicates that the bid strategy is not spending the optimal amount to achieve max conversion within the cost per action limit. If this is the case then it’s likely that higher ‘% cost err at max bid’ attribute correlates (with lower than 1.0) ‘% of optimal cost’ values.

If this sounds familiar then there’s a good chance your min and max bids are too narrow and the algorithm is not achieving optimal performance based on past auctions.

Difference from target position

When stipulating a position based target be sure to do this gradually for keywords that don’t have a wealth of historical data. Having either too narrow of too distant min and max bid limits could equally result in higher CPCs for auctions where the position could be lower, and lower position for auctions where the highest bid in the auction exceeds your stated bid ceiling.

The difference from target position attribute calculates the position difference between the target position and the actual position.

Listed above are a small handful of the measurements available within DS bid strategies which help to gauge bid strategy health and highlight any potential barriers to achieving optimal performance in line with your bid strategies targets.

Step 3 – Test and refine

In order for the tool to begin performing optimally, it is recommended that 4-6 weeks minimum is allowed before making changes unless of course poor performance of biddable items dictates that changes are made prior to this time.

In essence, the more time the tool is given to gather data, the more chance it has of achieving your goals.

For some advertisers, this could vary considerably depending on the amount of traffic being pushed through the platform. So if your client only receives 100 clicks a day then you’re going to have to be very patient.

Once the tool has gathered enough data then making changes using the measurements discussed above as your guide will be made much easier.

DoubleClick automation best practices

For client accounts containing hundreds of thousands, or even millions of keywords, as search manager the day to day management even at campaign level can be challenging.

No doubt about it, it’s encouraging to see accounts with abundant keyword data, it is in fact testament to search managers who have endured the pain of continuous expansion in the pursuit to deliver more results to meet ever growing client targets.

Introducing automated rules and scheduled alerts to your strategy can take away some of the heat and allow search managers to become less reliant on manual daily optimisation.

Automated alerts and rules

It’s a search manager’s worst nightmare. You leave the office on a Friday with everything in place for the weekend. You know your access is going to be limited over the weekend while you’re off visiting Aunt Brenda, and there’s a little voice in your head saying “I wonder if that page went live?” or “I wonder if that campaign ended?”

More often than not, it’s all perfectly well, but in the event of things not going so swimmingly over the weekend when new activity is going live then automated alerts are here to ease the anxiety.

For example, if one of the conversion pages on site stops working during developmental changes then setting up a simple low CvR% rule can help notify you before you frivolously throw more money on driving traffic to that all-important promotional page which is dead. For example, the rule here would be (CvR% < 0.5%: label = “PPC – Performance”). The alert would notify me that the account label ‘PPC – Performance’ has dipped below its average conversion rate and I need to take action.

The action could be either to pause all activity (or depending on the type of issue) enable some other activity which will remedy the poor performance.

If your primary KPI is volume of clicks but conversion is your secondary KPI, then switching from manual bidding to bid strategies could help you to achieve both. Example below:

A simple change of keyword state can be the difference in achieving your client targets and not. We’ve all had those weeks where we’ve been only a whisker away from being in the green.

Automated rules take away the worry of investing in keywords that aren’t working. Simply tell DS to pause keywords which do not meet a certain performance criteria. Example below:

The Google SERPs can be very volatile and often a keyword in top positions may cost more one week than the next, plus varying CvR% may result in said keyword being optimised down or even switched off.

Although we know that this keyword has the potential to perform well some weeks it’s important that we still remain within the clients target cost per conversion. So when this keyword is having a bad week we can automate to reduce bids and vice versa.

The below gives an example where the automated rule increases the max CPC using a combination of CvR% and click conditions:

So there you have it. A few handy automation’s that will save you time and leave you safe in the knowledge that your accounts are ticking over nicely while you enjoy a nice cup of tea with your Aunt Brenda this weekend.