Paid-search bid optimization comes in two flavors: rules-based and model-based. Within the broad realm of model-based optimization, you’ll find three common methods: global cluster-level modeling, local keyword-level modeling and global keyword-level modeling. Each has pros and cons and each offers various degrees of performance improvement.

If you’re not certain what method you’re using currently, or are considering changing what you’re doing, read on for an overview of these different optimization approaches and their relative strengths and weaknesses.

Model-Based Optimization

Global keyword level: Every keyword receives individual analysis and an individual bid so the entire portfolio achieves the stated goal. The approach is difficult to accomplish because the size and complexity of many paid-search programs requires literally millions of pricing decisions to be made each day. But a solution that does accomplish global keyword-level modeling is likely to be pure software and, therefore, fully automated.

Local keyword level: In effect, this is the approach advocated by Hal Varian, Google’s chief economist: bid each keyword separately based on the predicted value. Simplicity is the key because you don’t need to predict behavior across a range of bids – just bid a percent of the predicted value. However, with simplicity comes lower performance and limited settings. In a lot of cases, a local solution leaves money on the table. You can set a target but you can’t layer on multiple constraints.

Global cluster level: Here you still have a global optimization but models based on clusters of keywords are used to handle the sparse-data problem. Some vendors are actually a hybrid of this plus the keyword-level global approach — that is, they use keyword-level for head terms and clusters for the tail. Global cluster-level bidding tends to be stable with results that are repeatable. But what you gain in stability, you lose in performance and automation. Performance suffers because clusters ignore the fact that every keyword is unique and the value of the aggregated data is outweighed by the loss of that uniqueness. A variety of factors such as seasonal changes, expanded keyword lists and changes in product offerings can render clusters obsolete. When obsolescence occurs, statisticians are typically needed to manually tune the models, so cluster-based solutions are rarely pure software applications.

Rules-Based Optimization

The most common solution available, rules-based optimization is touted for being simple and easy to understand. For example, a rule might state, “If ROAS is less than 200 percent, lower bids by 10 percent.” However, when rules are layered upon rules, simplicity and understandability are quickly negated, making it difficult to understand what will happen to the bids. The big loser in rules-based optimization is performance. Rules-based systems are reactive, with pre-defined responses to certain situations. In a rules-based system, historical data is not considered because the situation drives the reaction. Because of their reactive nature, rules-based optimization can be very good at protecting your position, but playing defense rarely leads to optimal results.

For a more granular look at these approaches, download OptiMine’s white paper (no registration required) “Achieving the Gold Standard in Paid-Search Bid Optimization”.

@markpalony

eMarketer is out with a new study that shows that the biggest challenge facing companies, in terms of “Bid Data”, is “the time and manpower required to collect and analyze it.” The full chart of answers is below and it is almost a graphical representation of the throwing up of ones hands in a surrender to the issue at hand.

Now, we all understand the importance of data and we can all understand the frustration that comes when there is too much data to consider, but consider it we must. The key to success is how one goes about the task. For my purposes, I like to start at the most granular level possible – the DNA level, if you will – and build my understanding from there.

If you’re trying to make sense of, not to mention decisions based on, your entire digital marketing data set without first understanding the individual channels, you are setting yourself up for suboptimal performance. Put another way, it’s easier to take on the forest after you understand the individual trees.

An interesting study has been released by RevTrax. The study was conducted over a two year period and found that paid search had a significant impact on in-store sales. The summary, which you can download here, highlights some cool stats:

• The average click on a paid search ad generated approximately $15 of in-store revenue, with some
  merchants seeing as much as $28 of in-store revenue.
• Approximately 9% of clicks on a paid search ad generated an in-store sale, with some merchants seeing up
  to 26% of clicks on a paid search ad generating an in-store sale.
• Many merchants reported that 40-50% of customers acquired via the study were new

To conduct the study, RevTrax used the following process:

• a paid search advertisement was displayed to a consumer;
• the paid search ad led the consumer to a printable or mobile landing page displaying a coupon with a
  unique barcode; and
• the consumer redeemed the coupon inside a brick & mortar store.

Utilizing its unique barcode or promo code technology, each coupon was tracked back to the search activity that
drove engagement. RevTrax also utilized a variety of proprietary coupon security features to ensure clean
measurement of data only from the paid search channel.

Another interesting bit of info that’s not in the summary, but can be found on Search Engine Land is that, in the words of RevTrax COO Seth Sarelson:

As for PC vs Mobile, we’re working on a later study that will address these results, but most of what we’re looking at here is printable coupons from a PC.

Before I found this tidbit, I was ready to assert mobile as the primary driver of the results. After all, with the proliferation of smartphones and tablets, it would seem to be a  logical conclusion.

I look forward to the follow-up study.

@MarkPalony

Whether you’re new to paid search, or a seasoned veteran, “Achieving the Gold Standard in Paid-Search Bid Optimization” lays bare the truth: that all model-based optimization is not the same and the differences are significant in approach and results.

Download your copyof “Achieving the Gold Standard in Paid-Search Bid Optimization” today.

There are several ways to optimize your keyword bids, but they are not all equal. You’ll find a summary of them on the OptiMine website here and a more expansive look in our whitepaper, Achieving the Gold Standard in Paid-Search Bid Optimization.

Regardless of how you’re optimizing today, do so with an eye on the future making sure to track your conversions at the most granular level possible: the individual keyword. Even if you don’t use it today, having the data available will make for a smoother transition if you need it in the future.

• Most bidding software relies on large quantities of historical data to forecast an optimal bid. Those products will often cluster hundreds of similar keywords together in   order to gather enough data to generate a bid. By doing so, those technologies lose the “unique identity” of a keyword, Moar said. What distinguishes OptiMine is its ability to model keywords individually and do so with minimal statistics.

• OptiMine was founded in 2008 by Cooley and Doug Bryan, both of whom previously worked for data-mining company Kxen Inc., based in San Francisco. While there, they  often heard from customers who wanted a better bidding system. Cooley’s data-mining expertise was key to developing the math behind OptiMine, said Moar, the company’s CEO. Moar is a former CEO of Identix, a Minnetonka-based biometrics firm that was later sold to L-1 Identity Solutions.

• Earlier this year, the company landed a $3.6 million round of venture capital, which it plans to use to increase its staff and to draw awareness to the company. The round was led by Hummer Winblad Venture Partners    of San Francisco. Mark Gorenberg, a managing director at the firm, said OptiMine is well positioned in the paid-search market. He added that the round of financing should help the company gain further traction.“It sends a signal that they’re a major player in the marketplace. And it gives comfort to their customers that they’re here to stay,” he said. “The market is good for OptiMine. It’s a huge opportunity for the Twin Cities to get a good lead in optimization software.”

OptiMine has enjoyed great success in 2011 and we are very much looking forward to continuing the streak i n 2012.

ST. PAUL, Minn. – Oct. 24, 2011 – OptiMine Software has secured $3.6 million in a Series B round of funding, led by Hummer Winblad Venture Partners. OptiMine, which develops keyword bid optimization software for large paid-search advertising programs, will use the funding to make significant investments to expand its product development and sales capability, among other activities. Tests show OptiMine consistently improves paid-search advertiser returns by 25 percent or more.

OptiMine’s bid optimization software uses multivariate predictive analytics to manage keyword bidding for paid-search campaigns. Its patent-pending technology analyzes the performance of every keyword individually using 24 models and examining more than 100 variables to determine and automatically set daily keyword bids to maximize overall performance. OptiMine backs its software with a performance-based price guarantee.

“OptiMine has solved a problem that faces all large paid-search campaigns – how to make millions of complex pricing decisions every day in order to significantly improve and maximize performance,” said Hummer Winblad managing director Mark Gorenberg, who will join the OptiMine board. “We believe it is the first mover in a new generation of optimization analytics that will drive superior performance across the online advertising industry.”

OptiMine was founded in 2008 by experienced software executives with half a century of experience in using predictive analytics and data mining to improve marketing performance. It raised $1.1 million in Series A funding in 2010, led by Calumet Venture Fund of Madison, Wis. Calumet is increasing its OptiMine investment through participation in the B round. In September, Adobe announced a partnering agreement to add OptiMine’s keyword predictive analytics software to Adobe’s Online Marketing Suite.

“We are thrilled to have a firm with Hummer Winblad’s expertise and success investing in OptiMine,” said Jim Moar, CEO of OptiMine. “We look forward to moving faster to help more advertisers improve their paid-search advertising returns and to expand our product offering across the online advertising industry.”

About OptiMine Software

Based in St. Paul, Minn., OptiMine Software develops software that dramatically improves online advertising performance using more and better math to solve complex pricing problems.  For more information, visit www.optimine.com.

About Hummer Winblad Venture Partners

Hummer Winblad Venture Partners is a leading venture capital firm focused on software investing and manages over $1 billion in cumulative capital. Since Hummer Winblad Venture Partners’ inception in 1989 the firm has launched over 100 new software companies. For more information, visit www.humwin.com.

So, if you want to fail, read Bills post and find out the details behind Quinns sound (bad?) advice like:

1. Create a “Kitchen Sink” Ad Group
2. Use Broad Match Keywords
3. Use Irrelevant Keywords
4. Keep Promotions and Free Shipping a Secret
5. Use your home page as a paid search landing page

Follow these five steps and you, too, will find your campaigns on the paid-search ash heap.

Quinn’s list may be 100-level information, but it serves as a good reminder of two things. First, no matter how much of a paid search expert you are, the basics are always important. New people are entering the paid search world every day and having a sound what-not-to-do foundation will serve them well.

The Q&A comes from Josh Dreller and his In The Trenches column found in search engine land. Josh is a true expert in the industry who “gets” why auto-bid technology is necessary for paid search, but also understands the danger in the set-it-and-forget-it mentality that often accompanies its implementation.

In the interview, Josh gets Rob to break a difficult problem into its component parts and describe how they are put together in a way that benefits paid search advertisers. If you’re not sure what to look for in auto-bid technology, this primer will give you enough information to begin asking the right questions.

In it Rob talks about the four approaches auto-bid technologies use to optimize paid search bids: Global Keyword-Level, Local Keyword-Level, Global Cluster-Level, Rules-Based. You’ll also get a quick look at models vs. rules, keyword vs. clusters and global vs. local optimization.

If you’re managing large paid search campaigns automating the bidding process will drive efficiency, but the right auto-bidder will also drive performance improvements. Others may have been stopped short of finding the Holy Grail, but it does exist.

Read the interview and you’ll know what I mean.