Review LOB Link Distributions

OVERVIEW:

  • Review Link Distributions – Primary Language
  • Review Link Distributions – 2nd language
  • Review Link Distributions – 3rd language

PROCESS:

An important element of any SWXL update is to ensure that the link distributions across Lines of Business are in-line with the business objectives identified by the Specialist. These objectives are usually indicated / discussed in the kickoff meeting at the beginning of the project.

Specifically, we should take into consideration the relative % of links that are assigned to Hotels, Flights, and Vacation Packages, and compare them to less-profitable LOBs or those that are not as heavily targeted such as Cruises, Cars, and General.

While flights are not as profitable, they typically will have a fairly high relative % of links (and URLs and KWs) because there is typically a high count of Flights-related URLs in inventory (especially if the POS has a Flights CMS). Keyword counts tend to be fairly high as well, as there are a number of variations to the core Keyword “flights to _”.

The QA team should also take into consideration live versus update statistics, to see that growth across LOBs in terms of URLs, KWs and Links makes sense – did new pages (inventory) get added? Did the ratio of KW:URL change? Did the specialist add or delete KW variants?

Generate Simple-Names

  1. Ensure that ‘AtlasVersion’ and ‘ProductHotelsVersion’ are set correctly in the Config table.
  2. Create synonyms using build-db.pl:
    • all scripts under https://dev.myersmediagroup.com/mmgsvn/KMS/trunk/KWGeneration/trunk/db/synonym/gkms-atlas-references/
    • all scripts under https://dev.myersmediagroup.com/mmgsvn/KMS/trunk/KWGeneration/trunk/db/synonym/gkms-product-references/
  3. Build new simple_names tables with:
    https://dev.myersmediagroup.com/mmgsvn/KMS/trunk/KWGeneration/trunk/KWExplosion/preprocess/02-all.sh

Import Keyword Research Results

OVERVIEW:

After receiving root and research terms back from SEOClarity, we need to import the Rank & SV data into the POS database.

PROCESS:

Use: https://dev.myersmediagroup.com/mmgsvn/KMS/trunk/DataSources/STAT/script/seoclarity-import.pl
(This is also step 4 of eSV-Research-process.txt)

Then run: https://dev.myersmediagroup.com/mmgsvn/KMS/trunk/KWGeneration/trunk/KWExplosion/preprocess/03a-Updated_Latest_Data.sql
(This is also step 5A of eSV-Research-process.txt)

 

  • Import SEOClarity – Primary language
  • Import SEOClarity – 2nd language
  • Import SEOClarity – 3rd language

Review Template Distributions

After reviewing LOB Distributions, a deeper dive should be taken on the major templates. Often this is where you can see causes of LOB Link Distribution shifts when comparing Live vs Update data at the LOB level.

Specifically, look at the Template % of KWs and % of URLs in Groups 1 and 2, as well as template priority. Does the template priority make sense?

Look for changes (compare Live vs Update data) – has there been significant growth (or shrinkage) within a template of URLs, KWs or Links? If so, is there a reasonable cause for the change, such as new URLs coming on line, new KW variants being added (or deleted)?

Update Hotel Gross-Bookings

Primary Language Only

OVERVIEW:
Download and import the latest Hotel Gross Bookings (Hotel GB) data into the SWXL DB for use in URL and Keyword Generation processes.

– update GBV (gross bookings)
– update Transactions
– update Room Nights

You can find the latest Hotel GB for each POS under:
https://drive.google.com/open?id=0B2nosImlsF9HU1pvY05KdGRMODQ

Otherwise, we will have to download it from Expedia through their VPN or request it from Expedia.

DOWNLOADING & IMPORTING:

  1. Enable your Expedia VPN. Instructions for setting that up can be found here:
    https://docs.google.com/document/d/1w18zj4B49KUxuFshyYYMTM-cbPWYZsCqYTcy5Ixc9wo/edit?usp=sharing
  2. Go to http://hue/beeswax/. Log in as jmihalik / jmihalik.
  3. In the big query window, run the attached query.
    (Adjust the point of sale and date to be about one year ago)
  4. SELECT
     s.site_name,
     a.EXPE_LODG_PROPERTY_ID,
     a.LODG_PROPERTY_ID,
     a.LODG_PROPERTY_NAME,
     SUM(b.RM_CNT) AS ROOM_NIGHTS,
     SUM(b.RM_NIGHT_CNT) AS ROOM_NIGHT_CNT,
     SUM(b.GROSS_BKG_AMT_USD) as GBV
    FROM
     DM.LODG_RM_TRANS_FACT b
     INNER JOIN DM.LODG_PROPERTY_DIM a ON (b.LODG_PROPERTY_KEY=a.LODG_PROPERTY_KEY)
     INNER JOIN DM.site_dim s ON (s.tpid=b.tpid)
    WHERE
     b.trans_date_key >= '2016-08-23' -- about a year ago from when you are running the query to get 12 months of data
     and s.site_name = 'EXPEDIA.CA' -- site name selected from list below
    GROUP BY
     s.site_name,
     a.EXPE_LODG_PROPERTY_ID,
     a.LODG_PROPERTY_ID,
     a.LODG_PROPERTY_NAME

    Query should only take a few minutes.

  5. Save results as csv.
  6. Run the appropriate query from: https://dev.myersmediagroup.com/mmgsvn/KMS/trunk/KWGeneration/trunk/KWExplosion/preprocess/06a-Hotel_GB.pl