Search engine marketing process

Repeat after me. Search engine marketing (SEM) is not just a technical process.

Over the past 13 years of doing this internet marketing stuff I have seen many technical SEO approaches and it seems in Estonia that many web development agencies relate to search marketing challenges as purely technical issues. They are not.

There are far more things to consider when a company is looking to take FULL advantage of search channels.

So, in the continued spirit of sharing we open up (part of) a summary of our process which is based on successful experimentation, research and best practice in search marketing. We are happy to share this because we know that the devil is in the detail and the execution is tough and time consuming, whatever knowledge you start with.

This process is based on the STANDARD needs for a LARGE company and would be overkill for a small company.
If you have a small website then some of these steps should be eliminated.

Search engine marketing is a specialist skill – that’s our message and its something you can’t afford to get wrong.

People are searching for products and services like yours every day.
Each day you delay has an opportunity cost issue.
Each day you delay means missed sales and leads and more for your competitors.

STEP 1 is Identify the goals of your website/SEM campaign. Specify clear desirable actions e.g. sales, trials/leads, downloads, awareness building, opinion forming etc.

STEP 2 is to marry the goals against SEM types of i.e. Organic aka SEO aka Search Engine Optimisation or Paid e.g. Google Adwords.

To illustrate this, it is typical to use paid search for market awareness i.e. brand building or short-term promotions.
Opinion forming/informational goals are often best served with an organic approach.
Most activities benefit from a blend of both.

STEP 3 Measuring and benchmarking. We need to analyse the site and identify the processes/event chains/micro-conversionary steps that lead to a goal (online). Each cycle for the primary goals should be identified and statistics gathered to illustrate the conversion rates of: visitor-to-goal.

STEP 4 €€€€€. We need to identify the cash value of each goal e.g. you probably can already say how much a “free user” is worth based on the average conversion of free user to paying one. By working the maths backwards we can identify a value per visitor. This is key objective of this step. To get a cash value per visitor.

STEP 5 Defining the test. Before we get into analysis we have to decide what to analyse and agree a suitable test that will achieve the project’s goal.

The test should have the following characteristics: High impact (will impress stakeholders), easily measurable, not too difficult (so success is likely) and ‘doable’.

What’s important is to pick an important and ‘hot’ part of the site as a place to start.
Once we have targeted the “area” of the site we can do keyword research to identify terms that searchers actually use.

However we can be sure that there will be generic terms and also brand related terms.

NOTE: It is a possible sub-step to analyse ‘tough ranking positions’ and looks for gaps in the top 10. In other words if we can not reasonable expect to penetrate the top 10 then we need to think ‘paid’ not ‘organic’. Off and on page analysis of top sites for a given term can identify the likelihood of successful entry to the top ten list.

STEP 6 Site analysis. When we have our ‘candidate basket’ of search terms agreed (we can always change them later) then we need to identify appropriate landing pages within the site. If your website does not carry pages that provide relevant information for the searcher then the term must be eliminated. A test is probably not the right place to be requesting new content or tailored landing pages. We need to find pages that reinforce the searcher’s need.

Then we need to check the pages are actually indexed in our target search engines: Google, Neti, Yandex, Rambler, Yahoo, MSN etc.

Hopefully there are no indexing issues. If there are then we should either focus the test elsewhere or risk drawing negative attention to the need for structural change. Requesting site modifications is probably not a good test feature. Pain before gain 🙁

Assuming all is Ok we can check current rankings and develop a competitor matrix for the target terms using (in Estonian)

Then we check the site logs to identify existing traffic levels to see what

STEP 7 Maths. We can, using keyword research tools, make some gap analysis estimates about searcher demand and calculate the opportunity i.e. the difference between number of searches and the amount of traffic you actually receive.
NOTE: none of the keyword research tools are precise about numbers so there is an element of guesswork.

As we should already know the conversion ratios and ‘cash values’ for these keywords then we can say how much this increase is worth and therefore determine the financial value of the activity and therefore the ROI potential.

Research tools (for Estonian market) include: Google Keyword Research Tool and Neti’s (scroll down).

It is possible, if necessary, to do clever guesswork maths, to project potential traffic volume but frankly the results are too theoretical for my liking. It is possible though.
What’s better is just to calculate the range of potential search engine traffic increases.
We can also project potential rankings too if necessary.
By predicting the various factors we can estimate traffic increases for the test keywords.

Step 8 Scope.
Then we need to scope and divide the work (central v external/national) and assign roles to central / local / department teams.
Roles include: Strategy, targeting engines, keyword research, bid management for paid search, seo, tech modifications, tools selection, metrics.

· Create the business case for extension of work based on results
· Create FAQ for various stakeholders: national teams, brand managers, IT, finance, PR, legal, IA etc.
· Prepare presentations for each group
· Ultimately prepare exec level doc that answers the key questions.
· How much, how long, why, and what about the competition, risks, priority levels etc.


Loading Facebook Comments ...

There are 2 comments .

Share Your Thoughts!

Copyright ©2005-. All Rights Reserved.