Step Wise SEO

Using search engines to drive accountable sales via a fully eCommerce driven website is one of the most common and effective uses of the medium.

The use of search engine marketing for eCommerce sales is one of the most well documented as it allows for the greatest accuracy with measurement.

Through the use of accurate tracking and careful planning a series of measures can always be taken to ensure that the correct return or cost per sales is being achieved.

Submission type
This strategy requires all forms of submission type. As many websites are abundant in content, for example retailers and publishers, the representative keywords and engines available are vast.

The challenge in many respects is selecting the most appropriate to gain success.

Paid For Placement (PFP) - the paid for placement engines, such as Overture and Google AdWordsTM, are very important for a sales strategy.

With the largest inventory (i.e. number keywords) and strongest syndication (distribution network) these providers are the backbone of any campaign.

Organic search - as with any strategy or objective from search engines, the use of organic search is of paramount importance.

This low-cost opportunity presents the marketer with the possibility of optimizing a website to drive sales with low to minimum costs and high return on investment.

Paid For Inclusion (PFI) - PFI is well suited to sites that have large volumes of products to sell online.

PFI gives the opportunity to optimize more specific search variables such as SKU numbers, product descriptions, sizes, quantities etc, without having to select each and every keyword and submit via PFP

Vertical engines - engines such as Business.com or Industry Brains (US sites specifically) offer the marketer specific audiences.

For a strategy such as this it is advised that testing should be conducted to see the response in terms of ROAS.

An additional note is that vertical engines are typically more expensive than traditional PPC engines owing to a specific audience.

Engine selection
PFP First tier - these engines are the search providers that will drive the majority of sales. With the strongest syndication and largest keyword inventory the variations in keyword, creative and bidding strategy makes these engines very flexible to manipulate.

MSN featured sites will, if negotiated carefully, provide a flat rate CPC that could result in guaranteed top 3 placements for a long period of time.

PFP Second tier - many of these second-tier engines have a good probability of driving some very cost-effective sales.

Sites like FindWhat and MIVA both have strong enough distribution network and keyword inventory to make them viable propositions.

The other engines may also have the potential to drive sales; however, the likelihood is lower so testing is recommended.

Organic - natural listings are essential to any strategy. For a sales-orientated focus the organic listings possess the qualities to improve any search campaign.

If optimized correctly the traffic received from this type of search has no direct cost associated with it and any sale generates a very positive ROI.

With up to 66 per cent of the screen real estate being taken up by these listings, this type of search engine cannot be ignored.

Vertical engines - an option for any sales focus, these engines have good distribution into sites that are relevant to specific subjects, namely the B2B audience.

Success may be possible here if the product or service being sold is directly relevant to the searching audience.

Engines such as Business.com or Industry Brains will have to be tested as the specific nature of their listings carry a premium over many of the other paid search providers.

Shopping engines – an option but one that should be seriously considered. If the product is viable, appropriate categorization within the leading shopping engines, such as Kelkoo in the UK and Shopping.com in the USA, can result in strong return on investment.

Keyword and creative selection
Once the submission type and distribution network has been identified (i.e. the engines) the next variable is to determine the components that go into representing the company within the vast array of searchable variables.

In many respects this next section relates mainly to the engines where keyword selection is possible (i.e. PFP engines), however, keyword research is also extremely important for organic search.

Keyword selection
Branded terms - these terms are as important for a sales strategy as for any other type of objective.

Being terms the most likely to drive qualified visitors to a website compared with any specific product term, the approach to be taken with the research of branded terms is to ensure that every variation is covered in the initial campaign launch.

Generic terms - generic terms will drive large volumes of sales: the issue will be at what cost.

Generic terms will likely drive a large percentage of traffic to the site and many other advertisers are aware of this, resulting in increased CPC prices.

Additionally companies adopt different strategies: some might be utilizing search engines for branding or awareness and have more tolerance to higher CPC costs than a marketer looking to make cost-effective sales.

The recommendation with generic terms is to use them in moderation and measure their success against distinct measurement and control parameters.

Product terms - product terms are essential for a sales strategy. As Generic terms hold value in search velocity, product terms hold value in relevance and high return.

Product term characteristics consist of driving low volume traffic that has higher conversion likelihood.

In order to drive enough volume to make a difference to the impact of a search campaign, it is recommended that a large number (number determined below) of terms be used.

To find the ideal number of product terms really depends upon the mechanisms used to measure success.

However, the following process should be considered when researching these terms (this relates to PFP engines specifically:

• Research process: As a starting platform it is important to ensure that the core areas of business are covered so an approach could be to determine best selling products and research around them.

 Once done, expanding the product keyword list into other areas of the business could be a viable next step.

• Build from success: adding keywords is often an interactive process based upon the success of existing keywords.

If a keyword has proven successful based upon the parameters set for the campaign, use that particular word via the Research Process to build new phrases.

Add these to the keyword inventory and repeat. This type of expansion is often referred to as stemming.

• Test: It is important in any search campaign to test. In this instance it is important to nominate a budgetary amount that can be used for testing.

Once identified add keywords and test over a set period of time until poor and good performers are identified.

Poor performers can be discarded, good performers can be added to the keyword list and keywords with an indifferent performance can be tested again.

Match types
When selecting keywords the match type is key to success. If the wrong match type is applied to a keyword, the resulting traffic number could jeopardize the overall return of the campaign.

It is for this reason that user discretion is applied to the match types and should any broad or advanced match types be used, negative keywords should be carefully researched. As an initial guide the following should be observed.

• Branded terms: broad match types are acceptable for branded terms to cover all of the variations of a brand name.

• Generic terms: matching as specifically as possible is advised here unless heightened resource is provided, as the more broad or generic the match the higher the likelihood of irrelevant traffic that doesn’t covert well into sales and good return.

 Negative keywords are essential on generic terms as the multitude of searches possible is beyond the budget and aims of most marketing campaigns.

• Product terms: product terms, by the nature of the search characteristics (low traffic, specific search) held, can support a very broad (advanced) match type.

Creative development
For a sales approach the creative used must qualify a searcher as a potential customer and be compelling enough, with a call to action.

Branding creatives are those that continue a theme of maintain continuity with other branding or marketing activities.

In a sales strategy the branded terms can be used in multiple ways to increase awareness of a company’s particular brand, sale or product range.

This ties in very well with the specific creative listed below and as such the direct branding creatives are an option for this particular approach.

General creatives are those that represent the core benefits, offerings or values of a company but are designed to reinforce the fact that the searcher has come to the right place or found the right company.

In many instances search results are dominated with listings that might not be the actual site being looked for.

Affiliates, resellers, and competitors all might be buying or optimizing against a particular term and in most instances if the keywords being used are generic the likelihood of the maximum number of other bidders is high.

For this reason if someone is searching for ‘digital cameras’ and a listing from Canon appears with ‘digital camera’ written within the title and description the searcher is reassured that a trusted brand such as Canon is also presenting a link to ‘digital camera:

Specific creatives are those that utilize the titles and descriptions to represent a specific offer or message related to the keyword or searcher’s interest.

In many respects a specific offer creative is often used when a seasonal variable or event gives a company a short-term competitive advantage.

Measurement/control parameters
Once all of the actionable variables (i.e. engines, keywords, creatives) have been highlighted the next stage is to get an understanding of how to measure the performance against the objective.

In this instance it is sales so the following control parameters are used as well as the measurement thereof.

Control parameters
The most important variable here is to determine what the success metrics are and how they are to be measured.

In a sales strategy the perfect solution would be to make all traffic to a website accountable.

Understandably, many sales are generated offline from online traffic but to determine success online the following parameters should be considered.

• Return On Investment (ROI) - A simple calculation of revenue divided by cost, is one of the most effective ways of determining the performance of a search engine marketing campaign.

If a predetermined return is needed, a search campaign can be controlled to that particular level.

• Return On Advertising Spend (ROAS) - this calculation is the same as an ROI calculation except that the marketer will have determined what return is needed on advertising expenditure rather than advertising plus all additional costs of sale.

• Cost Per Order (CPO) - a CPO measurement takes into account the number of conversions (sales) generated by an advertising medium.

If the cost of advertising (in the case of search cost of clicks) is divided by the number of sales, the CPO is determined. Many companies have a range for a CPO and optimize within that range.

Measurement
With the identification of the control parameters the focus must now be to look at how to measure each of them to gauge success or failure.

Using measurements emanating back from the search engines and parameters gathered by tracking and analysis tools the following apply to the tracking of a sales (eCommerce) strategy.

Measurement variables
The essential criterion for measuring an eCommerce strategy is to be aware of all the parameters that determine success.

The most important variables to use are traffic, cost of traffic, conversions (generated sales) and if possible value of sale (which can be substituted for an average order value [AOV]). There are three means of obtaining this information.

1. Search engines: many of the engines, such as Overture and Google AdWordsTM, provide the search marketer with the option of taking all of the abovementioned data.

 This is free of charge in most cases or can be obtained with a small premium. The disadvantage with this type of tracking is that despite the information being free it is only available for a selection of engines and will not give a complete picture of success/failings.

2. Tracking tools: Please click here to learn about the detailed description of the tracking tools available to track the entire search campaign.

The advantage of a tracking tool is that all search submission types (PFI, Shopping Engines, PFP, Organic) can be measured in one location and, by analysis of performance, compared type for type.

The disadvantage is cost and resources required to manage such a tool.

3. Homegrown systems: many companies have devised analysis tools that allow marketing departments to track the performance of a search campaign using the website’s existing logs.

By utilizing a code (such as an affiliate ID code) a company can identify the source of traffic and watch the progression of said traffic through the site.

If these data are compared with the data provided by the engines all of the variables for control are covered.

Detailed measurement
It is also important to ensure that measurement for an eCommerce strategy is as granular as possible.

Keywords should be measured individually so for some search engines, such as Google AdWordsTM or FindWhat that allow you to measure groups of keywords together these options should be ignored in favor of the singular keyword tracking.

For sake of reporting, keywords can be grouped together and summarized but in an ideal scenario keywords, engines and sales must all be recorded in minute detail.

Indirect conversions (deferred sales)
In addition, it is important to appreciate conversions both directly and indirectly. In other words it is vitally important to recognize that users may return to the site after a few hours, days, or even weeks, to conclude a transaction.

Measurement calculations
The measurement parameters for success will be marketer defined according to the business parameters of the company.

For many companies search is measured differently from other marketing activities, as it is accountable (i.e. revenue over costs).

Management and optimization
The final stage in working toward a particular objective is to understand the way to manage the campaign and to optimize the performance if it isn’t as good as expected.

It is important in understanding how to make the most of search engines and generating sales.

PFP management
The variables that require management are quite simple and represent the core foundations of a strategy around achieving sales via the search engines. The following are the said variables.

• Presence - ensuring that search listings appear under search terms that are most likely to prompt a click and an eventual conversion.

• Quality - ensuring that search listings that appear are converting the searcher into a visitor to a website.

• Cost effectiveness - making sure that all search campaigns are running the desired business goals set for the particular medium and objective (i.e. meeting an ROI goal).

Managing a campaign for sales (eCommerce) consists of two distinct methodologies, as with the branding and awareness campaign. These are:

 (1) Technology and
(2) Manual management

Technology ‘bid management’
With an effective sales campaign working with all types of keyword it is essential to use some type of bidding technology to aid in driving success.

The following stepped process should be considered when managing a campaign through a bid management technology.

Step 1 Research - It is critical to research the right keywords to use in a campaign and more importantly to apply bidding to each of them.

Elements such as bid price, keyword velocity and estimated conversion rate can be applied to terms and the price and position needed to sustain an effective campaign.

Step 2 Selecting keywords - keywords researched must be selected both by their singular ability to achieve the goal and as a portfolio.

Different bidding rules or practices must be applied to keywords with different capabilities.

Keywords that are generic, for example, may have stricter CPC limitations as they hold most risk in using a large percentage of a budget.

Brand terms, for example, may require top placements to ensure a brand-related search is always catered for within the top syndicated positions.

The keyword classification must be done based upon the research mentioned in point number I so that accurate keywords can be selected to achieve goals.

In applying the conversion principles, keywords should also be selected based upon conversion potential.

This applies to any kind of keyword management; however, applying the same conversion rate to a generic word as to a specific word may produce inaccurate projections.

When selecting the keywords to use and the bid price range or position range at whatever time of day, based upon whether the research demonstrates the keyword performing at the desired return level using averages.

Averages must be estimated in the initial setup of a campaign; then, as more realistic data are passed back from tracking tools, the averages must be replaced with actuals.

Step 3 Implement - once the keywords are selected it is essential to produce the most appropriate bidding rules for each term.

Selected terms that can justify higher positions must have rules different from those that hold more risk.

Use the selected keywords to apply bidding rules and implement the terms on an hourly basis.

In many instances bid rules must be applied to singular terms, however, in many instances, the same bid rule or bid management technique can be applied to more than one keyword.

As many of the terms indicated in Step 2 don’t hold a keyword velocity warranting concern, these can be grouped together and a ‘universal’ bid rule applied.

Some bid management tools are more advanced than others. The more sophisticated tools allow for market conditions and the performance of the keywords to determine the best bidding rules.

These systems take much of the guesswork out of bid management, as trend-based analysis determines the most appropriate time of day, day or week, by keyword, to place bids and listings to ensure maximum coverage within the PFP search engines.

Manual management
Manual management is defined as the control of a campaign using limited additional technology.

The focus is to find work around mechanisms on the words and phrases being used to ensure that a campaign’s performance is maintained at all times.

Without a bid management technology this is a very difficult and challenging task, however, it is not impossible.

The way to master this is to focus on the tools and mechanisms provided within the industry that are freely available for all to use. In this instance these are the bidding vehicles provided by the PFP engines themselves.

The effectiveness of these mechanisms is in fact determined by the amount of resource allocated.

The question becomes that of frequency and updating techniques. Many companies choose to use different ways to overcome problematical resource issues.

One of the worst ways to ensure a top spot is to purchase the number 1 phrase using auto bidding, and to place a very high maximum bid but, as highlighted, this isn’t one of the better ways to go about it. Other methods more successful than this have been employed, as follows.

• Manage keywords sorted by importance: if the keywords are sorted by importance and relevance to a company’s products, services or branding exercises, more emphasis could be placed upon these words only and the remaining terms updated less often.

• Manage keywords sorted by cost: for the more fiscally challenged campaigns the management of keywords by price is critical.

In sorting keywords by those that have a high likelihood to run away with a campaign, these should be managed more carefully and more often than those that don’t have as high a probability of doing so.

This rationale is particularly important for the management of Google AdWordsTM and in reference to organizing campaigns to get the most from each keyword based upon a budget.

• Manage keywords sorted by performance: As a search engine marketing campaign matures, information becomes available from many sources on the performance of each keyword.

 If this is applied to a conversion parameter, performance can be determined. Whereas a bid management tool allows for hourly bid management based upon performance, the manual mechanism is a great deal more crude – when managing on an ‘as and when’ basis, analysis is only available at intermittent times.

 The latter form of management does give an element of control, however, as keywords that consistently perform well, related to sales, can be isolated and managed separately.

PFI/Feed management
PFI/Feed management is something that is often neglected by companies. The resource burden of the PFP engines nearly always outweighs the importance of managing feeds for engines such as Overture Site Match xChangeTM.

As with any listing that appears within any engine the parameters for management should be as follows.

• Ensure relevance - when preparing a feed it is essential to ensure the listings are relevant to the subject matter of the landing page.

The title, description, and optimized content (such as the short and long description of the product in question) all possess a similar relevancy and have a theme.

• Have One to One (1:1) ratio of listing to product (landing page URL) - as a prerequisite to using a PFI engine it is important to maintain a 1: 1 ratio of listing to product URL.

 In other words if the site has a listing for one particular product, there can’t be two listings in a feed pointing to that same product.

 This is important for management as updated feeds will potentially be processed faster when the feed is both relevant and has this ratio.

• Use strong creative - like any type of listing in any search engine the searcher is expected to click on the link if it is deemed to match the search query.
PFI listings should have all of the properties for compelling the searcher to click. Please click here for more data on writing strong creative.

It is also important to remember that in many instances PFP is most effective when managed in bulk, i.e. when an entire catalogue of products is submitted. Overture Site Match xChangeTM will only take listings of this nature.

Optimization
In many respects, using the search engines to generate sales requires constant optimization as many of the variables, such as keywords and market conditions (bid prices) as well as demand for terms and seasonality, change on a constant basis.

Many of the problems that are commonly associated with a strategy designed to drive sales and some suggested optimization techniques are presented for review.

The optimization of a strategy to generate sales is a dynamic and fluid process that must be repeated on a constant basis.

Market conditions change on a constant basis, both by the nature of the competition and by demand from the searcher, and, as such, constant manipulation of the criteria that determine success is required, especially in the case of PFP engines, on an hourly basis.

When using the optimization methods, it is important to ensure that enough of a timeline has passed to enable assessment to be done.

Optimization based upon hourly or even daily performance may not provide a size of sample large enough to enable determination of either success, or (as above) failings that need correction.

Table 1.

Sales driven strategy problems
PFP Engine Problem Optimization Technique
High Traffic, Low Sales Volume Scenario A: The most common problem here is that the keywords being used aren’t as relevant as need be.

If traffic is being received, there is interest but upon arrival on the site it is not finding exactly what is required and leaving without converting.

Revise the keyword list to weed out terms that are driving large volumes of traffic that isn’t converting.

Scenario B: The keywords being used are in positions too prominent within the engines.

Number 1 isn’t always the best option as in many instances the user may click without first determining if this is the right search listing to solve the query.

Attempt to lower the bid prices to lower traffic and attract more qualified users
High Sales, Low ROI Scenario A: The sales that are being generated are doing so at a high CPC cost.

Revisit the keywords and study the generic terms to determine if the cost to benefit ratio is working on these terms.

Using bid management attempt to control exposure of these keywords to the most optimum times of day that will generate sales.

Pay particular attention to the terms that are driving traffic and no sales

Scenario B: The products being sold cannot justify the CPC charges.

Study the CPC price in conjunction with AOV to determine if CPC prices are too high
High ROI, Low Sales Volume The keywords selected are too specific and the products aren’t getting enough exposure.

Revisit the keyword list and increase the number of generic terms and product terms to cover more of the searching audience.
100% Sales from Branded Terms The branded terms are masking the performance of the other terms within the keyword portfolio.

Revisit the categorization of the keywords and study the performance of the generic and product terms.

In most scenarios with a popular brand up to 90% of sales can be generated by branded terms.

However, in analyzing the performance of each grouping a more accurate portrayal of performance can be determined
High Traffic, Low Sales Volume The feed may contain too many products that aren’t converting or have more informational qualities (i.e. the listings show up when a searcher is looking for information on a product with no intention to buy).

Scan the feed for SKUs that are driving traffic and no sales and remove them from the feed
High Sales, Low ROI The products being sold aren’t justifying the CPC of the rate card.

As the CPC isn’t as flexible as the PFP engines the products must be optimized to a level whereby at the site conversion rate after a sample number of clicks (say 100) the potential sale can justify the costs.

Remove terms that a) may drive too much traffic or b) as per the example below, can’t justify the cost.
High ROI, Low Sales Volume Scenario A: The feed is too specific. Attempt to increase the number of products within the feed

Scenario B: Only a select number of keywords are showing up. Have the feed analyzed for possible improvement

Organic (natural search) engines
One of the strongest ways to increase sales at a very strong ROI is to adopt organic search engines, such as Google, Yahoo! and MSN.

In many instances these mechanisms have low recurring costs, which are in actuality based upon the commitment from an organization to constant improvement, and can be very strong in driving top placements, under a company’s brand name in particular.

It is therefore one of the most important aspects of a search strategy designed around sales, however, is a great deal more complicated than that of paid search. For this reason the following should be utilized.

• Research when either (1) employing an outside firm to optimize a website or (2) placing a request for web designers or technical departments to implement websit changes, it is important to ensure that keywords that will represent the brand and company are used within the optimization process.

Above all else it is important to be listed organically against these terms alone as many searchers still to this day utilize search engines even to find company websites rather than typing in a website address.

• Measurement: for a sales strategy it is recommended that rankings are used as the control and measurement variable and that sales from organic traffic are measured by recording the data from log file analysis against conversion points within the website.

 While adjustment and optimization of words and phrases based upon success is difficult it is possible to record total sales and then measure additional organic optimization.

As organic listings aren’t as dynamic as paid listings, the optimization and control aren’t as fluid.

For this reason when optimizing organically it is important to accumulate a history or presence within the engines and build upon this.

The constant optimization necessary for organic search is performed via paying attention to detail regarding the keywords and phrases that are most important to a business and increasing relevancy, via code changes and link architecture, to ensure long-term presence.

Keywords: Vertical engines, Organic, Shopping engines, Keyword, creative, Branded terms, Generic terms, traffic, CPC, Product terms, term, search campaign, Match types, broad match, advanced match, negative keywords, Branding creatives, Cost Per Order, eCommerce, traffic, conversions, tracking, bid management, PFI management, Feed management, optimization,