Hi Friends,

Even as I launch this today ( my 80th Birthday ), I realize that there is yet so much to say and do. There is just no time to look back, no time to wonder,"Will anyone read these pages?"

With regards,
Hemen Parekh
27 June 2013

Now as I approach my 90th birthday ( 27 June 2023 ) , I invite you to visit my Digital Avatar ( www.hemenparekh.ai ) – and continue chatting with me , even when I am no more here physically

Saturday, 13 September 2008

RESUME SEARCH BENCH MARK METHOD

Pl. read my note “Benchmark” (GumClone) on the whiteboard next to my table dt. 14-01-06
(This note is in file (B) Post Launch)

From a “benchmarked” resume, we are first extracting keywords & creating a “KEYPRINT” (like faceprint mentioned above). Then we are trying to compare/match this benchmark “Keyprint” with all the other “Keyprints” (of all the other resumes) in our database.

Q:
Can “Keyprints” of all the resumes be extracted in advance & kept ready (for Comparison/match-index purpose), rather than extracting the keywords on-the-fly at that very moment?
Will advance creation of “Keyprints” save time & speed up search?

28/12/06


Abhi
Rahul

“BENCH MARK”

▶ See enclosed UI

▶ If this application/feature was already up-and-running on IndiaRecruiter.net, then we could have found Abhi’s replacement, real fast!

▶ Imagine

  • no resume search (traditional at least)

  • no filling-in “Search Parameters”

  • no worry about “Industry–Function” etc.

  • no need for structured resume database

▶ Simply “keyword-based” “Match-Index”
(probs/weightages taken from our GumMine)

▶ “Bench-Mark” resume sum total of weightage, taken as 100.

▶ In the process, every time any HR mgr uses “BenchMark” feature, we gain one latest/up-to-date resume! It goes into our database.

▶ HR mgrs may even use this feature even when there is no “resignation/need for replacement”!


Here is the extracted text from the image (Page 2):


In the enclosed UI, you will notice that
in the centre-column (Matched Resumes Found), all those found are scoring less than 100 (arranged in descending order of score).

Reason?
Simple.
We have treated the “Benchmark” resume as 100 (sum total of weightages).

But it is possible that the “Raw score” of the “Benchmark” resume (let us say) is 40.

Now, if the matchmaker finds matching resumes as follows:

Raw Score Match Index
60 150
50 125
40 (Benchmark) 100
30 75
20 50

Apparently BETTER resumes (above benchmark)
Apparently INFERIOR resumes (below benchmark)


Would you like me to now extract Page 3 ("If we can do this...") as well?


Here is the extracted text from Page 3:


If we can do this, things become really "EXCITING" for the HR manager!

Now, he thinks/believes, he has found/come across SUPERIOR candidates, who are even better than the benchmarked candidate/executive!

So there is great temptation to click open & read all those which are scoring more than 100.
(Rs. 1/= per click revenue for our partners!)

He may even click open a few which are scoring 95/96/85 etc.
Because, he may suspect that those scoring >100 may be highly placed / getting much better salary and therefore less likely to join his company.

(It is like small companies cannot attract IIM-A graduates, even to appear for interview!)

So, some more Rs. 1/= !


A match-index tabulation, with scores ranging from (say) 50 to (say) 150 — as shown on pg: 2, has another important use for the HR mgr.


Would you like the content compiled into a Word document next?


It helps answer following question of a HR mgr:

“Who are other executives in the market, who are comparable to my own executive? And, what kind of salaries are they getting?”

Are most of the “matched” executives better-paid than my own comparable executive? If that is the case, we are likely to lose our executives fast — may be to our competitors!

Apparently our “Compensation-Structure” needs some urgent/upward revision! — if we want to bring down our churn-rate / our attrition-rate.

So, even when a HR Mgr. is not actively looking for a replacement (against a resignation); he will still indulge in this “BENCH-MARKING” exercise, every now and then, to keep his hand on the PULSE of the demand/supply situation.

To keep a track of how competitors are compensating different kinds of executives.

To tweak his own compensation structure.


Of course, we must not show the “RAW SCORE” column. How we have arrived at the “Match-Index” must remain a MYSTERY/SECRET.

But,
may be, next to the “Match-Index” column, we can display a “Current Gross Compensation” column as follows:

Match IndexCurrent Gross Ann. Comp. (Rs. L/year)
1504.8
1252.8
1003.2
752.6
503.2

Obviously, candidates having
→ higher score
→ lower CTC

are of great interest to the HR mgr.
He may even want to interview them, irrespective of a “vacancy” — kind of “standby” measure.
With “salary” column, we are bound to generate many more “clicks”!


We can even add a twist, called
"REFINE SEARCH".

(Visual sketch of interface showing "Benchmark Resume" with PEN 12345, Match Index scale 150–50, and a "Viewed Resume" box with PEN 12345 at 75.)

Now add a feature, which permits/enables a HR mgr. to transfer any resume which is being “viewed” to “Benchmark” box, thereby making that resume as “Benchmark” resume.

And the process starts all over again, taking PEN 12345 as new benchmark.
And a new set of “Match-Index” column will appear, tempting HR mgr to start clicking all over again!
This is likely to become a HIT with the HR mgrs — making them go into endless “LOOPS” of refinement after refinement! Now we become a true:

  • Google of Online Recruitment

  • Google of Resume-Search

Now all that we need to do is to simply upload all the unstructured/text resumes which we get everyday. Simple as that.


Rajeev  

cc: Abhi / cc: Vikram / cc: Rahul  

03-02-06


Benchmarking — is a many-splendoured thing!


▶ See enclosed tabulations  

▶ This should be our attempt to leverage  

   • 1 year of effort  

   • Lakhs of rupees spent in developing GuruMine  


▶ These features could put us way ahead of Monster/Naukri  


Let us do


Also find enclosed, a list of Modules to be developed. Let us discuss when we meet tomorrow.

[Signature]


03/02/06

Modules to be Developed         | Person Responsible    | Time-schedule  
----------------------------------------------------------------------------------
1) Job Search - Anchor          | Rajeev                 | HIGH  
2) Job Post - Anchor            |                        | HIGH  
3) Resume Search (Prospect)     |                        |  
----------------------------------------------------------------------------------
4) Resume Courier               | Rahul                  | LOW  
5) Don’t Waste Your Time        |                        | LOW  
6) Recharge your Prepaid A/c    |                        | HIGH  
----------------------------------------------------------------------------------
7) Admin Tool                   | Vikram / Rahul         |  
8) Resume Search                | Vikram                 | HIGH (All 3 tools)  
   Benchmark Method  
9) Make Fun of it (Blog)        | Rahul                  | LOW  
10) Processing 2 lakh resumes   |                        | MEDIUM  
    using GuruMine  
11) Webpage Harvester           |                        | LOW  
    Subject Classification      |                        | LOW





















No comments:

Post a Comment