Notes: SALARY PROFILES

Pay check across jobs will be a breeze soon

Dutch firm Wage Indicator has begun compiling data in India

DNA, 20/12/06

Rahul – Saurabh – Sunit – Yogesh – Sonal

All of you should visit this site to check out how it works. What kind of details does it ask you to fill in online? Is it too much? Is it too complicated?
What do “survey participants” get in return for their time/data? How often can they come back to check?
In what respect is this superior to our own “Salary Profile”?

In course of time, we have plans to provide apple-to-apple salary comparisons for co-professionals who are truly “co-professionals” in the sense of belonging to:

Same Industry
Same Function
Same Desig. Level → Phase II (when we have large enough database)
Same Age range
Same Exp. yrs. range
Same City/Location → Phase III

(Signed, 20/12/06)

Workforce Profile (Express Computer)

Age distribution:
- 24–28 years → 90%
- 29–35 years → 10%
Academic Profile:
- Graduate Engineers → 75%
- B.Tech → 15%
- M.Tech and MCA → 10%

Global Recruiter V2

We have provided fields for Educational Qualification in Resume Form – so, after we have enough registrations, developing such profiles should be easy.

These need not be developed dynamically online. Could be done offline, once-a-week.

(Signed, 19/09/06)

WWW.Salary.com

16-07-04

STEP #1

Search is to be conducted by a JOB CATEGORY and CITY/STATE.

JOB CATEGORY

Most are Industry-Names
A few are Function-Names, such as:
→ Accounting
→ Admin/Support/Clerical
→ Customer Service
→ Human Resources
→ Logistics
→ Manufacturing
→ Marketing
→ Materials Management
→ Planning
→ Purchasing
→ Sales
→ Training

A Designation Level is also used, e.g.:

Entry-Level
Executive/Mgmt
IT-Manager
Skilled & Trades (3)

This mixing-up of:

Industry Names
Function Names
Designation Levels (or Actual Designations)

is quite confusing!

We will need to decide how are WE going to present salary-graphs – based on what Search-Criteria.

STEP #2

Display shows:

Location: Birmingham/AL (Alabama – state)
Job Category: Purchasing

Now it asks:

“Select the Job-title whose description is closest to yours.”

[Example dropdown]

Asst. Merchandise Buyer
Buyer I
Buyer II

Options:

Search by Keywords
Read Job Description

Display shows – BASE PAY ONLY

Graph:

25th Percentile = $32,818
50th Percentile = $36,938
75th Percentile = $41,133

Table Example (Buyer I)

Location	25th Percentile	50th Percentile	75th Percentile
Birmingham, AL (city)	$32,818	$36,938	$41,133
The United States (overall)	$33,903	$38,159	$42,492

Key Features of Salary.com

1200 unique jobs
4000 job-titles
Data covering > 1.3 million individuals
Data covering > 5000 companies
“Salary Surveys” purchased from multiple sources
A huge team of researchers & consultants
“Employer Reported” data
Geographic salary equivalent factor
Lot of manual work (matching job descriptions to most applicable jobs from each data source)
A proprietary mathematical model that takes into account:
- Statistical reliability of each datapoint
- Effective date of data reported
- Industry
- Geography
- Company size
Database updated “every month”
Data on typical cash compensation
Does NOT use info from:
- Individual site users
- Placement agencies
- Job postings
Weightage given to geographical area
(e.g., San Juan → 75% of National Average, Fairbanks → 124%, Toledo (Ohio) → 100%)
Free Data vs. Premium Data

Where (in what respect) will WWR differ from Salary.com, as far as salary data is concerned:

Our data compilation/analysis/extraction/presentation (entire datamining) will be completely automated – without any human intervention!
We will have no expert/consultant/support staff for this work.
There will be NO purchase of salary surveys from any source/any time.
The only basis (for database creation/analysis/presentation etc.) will be salary details furnished by the candidates in the “Career History” section of their resumes, as & when posted on WWR.

We will NOT try to verify/validate this data.
We will not even try to define “salary/annum” – as to what it includes & what it excludes.
For salary-graphs, we will use only the current job / last job (in case unemployed) salary because we can link this with the candidate’s current posting location (city).

For his previous jobs (even last job), we have no clue in which city he held that previous/last job! Hence, “city-wise” analysis will not be possible.

CITY-WISE analysis will be possible ONLY for “current jobs.”

We, too, will provide 2 types of “analysis” to candidates as well as corporates:

Minimum/barebone analytics → FREE
Premium analytics → PAID

FREE analytics will, hopefully:

Attract/motivate the jobseekers to provide their current/accurate salary data (by returning frequently and editing their Career History section as soon as there is any change). Every time they “edit,” they will get to see/view their latest “standing” amongst peers (those belonging to same “function” and having same “designation-level”).
Attract/motivate the Employers (the job-advertisers) to insist on jobseekers to submit their resumes only through WWJ. Employers will get to see/view the salary details (graphical) as well—only in the ImageBuilder version of resumes.

This analytics will enable an employer to compare the applicant’s salary not only amongst his peers but also as compared to similar persons already employed by the employer in his own organization. This would help an HR Manager to quickly “zero-in” onto those few applicants who “meet the employer’s salary-criteria.”

Whereas “FREE” analytics will help us expand our database of candidates and corporates, for making money we will need to offer PREMIUM analytics to both parties.

We may conceivably deliver the PREMIUM salary-analytics as follows:

To Jobseekers:
As a PULL service (MMS) through mobile-service providers on a revenue-sharing basis, through the WWJ platform.

This arrangement will take care of our “collection mechanism.” We just cannot (and must not) try to put into place any revenue-collection mechanism on our own, from millions of jobseekers. That would be madness!
To Employers:
As a “logon & search” service from our RecruiteGuru.com platform, on a pay-per-use basis (as a web service), to those corporates who become members of RecruiteGuru by paying an advance (cheque/credit card, etc.).

This will be a PREPAID service, like all other services from RecruiteGuru platform.

In light of the preceding observations, it is obvious that databases on WWJ & Reemtium will be synchronised in real time. There is no question of "every month" update, as in Salary.com.

As far as MATHEMATICAL MODEL is concerned, unlike Salary.com, we should not try to develop any proprietary model of our own.

We must use STANDARD, SIMPLE, STATISTICAL MODELS/FORMULAS, freely available on the Net – something which a great majority of users can understand with ease.

We should not try to inject/introduce our own "SUBJECTIVITY" by adding arguable WEIGHTAGES.

Users may not understand our "Logic/interpretations" and even if they do, they may not agree! So, why complicate matters?

People (users) may not "LIKE" raw-data but at least, they cannot "contest/disagree" with the same!

What kind of "ANALYTICS" (Salary-Comparisons) can we think of?

(For the moment, forgetting which is FREE and which is PREMIUM):

Industry-Wise
Corporates and candidates want to know what are salary-trends/values in the same Industry in which they are employed.

Having selected their own "Industry", within that Industry they may wish to further narrow down their "comparison" to executives belonging to a specific FUNCTION.

Within that FUNCTION they may wish to further narrow down to executives, holding/having same DESIGNATION-LEVEL.


Step #1 ↓   Industry  
Step #2 ↓   Function  
Step #3 ↓   Designation Level  Step #4 ↓   City

A candidate (or a Corporate) should be able to re-arrange the "narrow-down" process in any order/sequence of above-shown four steps. He may not even want to go thru all the 4 steps. He may want to see the "result display" after the first 2 steps only.

But he will have to “pay,” every time he clicks DISPLAY button!

He should be free to click DISPLAY button, after any step, in any order, and, as often as he wants!

We want him to “explore” (like a child) by clicking on thousands of possible permutations & combinations possible. Idea is to “amaze” him with graph after graph!

The more he clicks, the more we earn!

Of course, all these clicking will be only possible on Recrutiguru.com and therefore only for corporates.

As far as jobseekers are concerned, even if, technically, we could create similar feature on WWJ, we must not do that because it is simply impossible for us to handle millions of PREPAID accounts of jobseekers on WWJ—and keep track of “Balances”! Logistics will be insurmountable!

For jobseekers, we will only provide SMS (or MMS graph) on their mobiles, as a PULL service.

For each jobseeker registered on WWJ, we know:

Name / PEN
Current City
Current Salary
Function
Industry (we must in course of time allot “Ind” to each resume)
Design. Level

So, whenever on his mobile, he sends our server an SMS (thru his MSP):

SAL SMS 2580 (To receive Salary Comparison as an SMS)
OR
SAL MMS 2580 (To receive Salary Comparison as a graph).

Our server will return on his mobile:

Ajay

Function: Sales

Design. Level: Manager

Industry: Pharma

Population: 4392

Highest Salary: 2.6 L

Lowest Salary: 1.1 L

Median Salary: 1.8 L

Your Salary: 2.2 L

Your Percentile: 80

Kartavya

NOT URGENT – PROFILES

20/03/03

WebService / Population Profile

A candidate’s or an existing employee’s FUNCTIONAL EXPOSURE PROFILE

Can be compared with:

(A) Overall TOTAL POPULATION ("Compare/Total" option)

(B) Specific SUBSET POPULATION ("Compare/Peers" option)

Under (B), a person can be compared with other executives who are:

of same “Edu. Level”
of same Exp. (yrs)
of same “Design. Level”
of same “Industry”
of same Age
etc.

For definition of “same Age” I suggest we take all persons born in:

Preceding year (as person being compared)
Succeeding year

This means, if we want to compare a person born in 1972 – then we take Subset Population = all persons born between 01-01-1971 → 01-01-1973 (i.e., a 3 calendar yrs).

This way, we specify one class interval. It will always include persons of “same known birth year” and those just older/younger by 1 year.

No matter, since we are not adding-up subsets to arrive at “Total Population”.

Kartavya

12-03-03

Salary / Career Progression Profile

Along with my note dt. 10/3 I had enclosed 2 forms for capturing

(a) Salary data

(b) Designation data.

Whereas the form for "Desig.-Level" data seems fairly simple & easy-to-use (simple radio-buttons), the form for capturing "Salary" details would require a lot of accurate typing by jobseekers. No one would have that much patience!

I suggest following simplified form:

Year	Gross Annual Salary (Cost-To-Company)
Rs. Lakhs/year (Rs. 00,000)

Example:

1995 → Two digit max [ 9 ][ 9 ] . Decimal [ 9 ] Single digit

Assumptions:

▶ One decimal point is good enough accuracy.

▶ We don’t expect candidates getting Rs.100+ lakh/year to send-in resumes!

With only 3 digits to be typed, it is quite easy.

We could even consider 2 separate “dropdowns”, one each for:

Integers (having 0 to 99)
Decimals ("0 to 9")

We may not ask candidate to click Submit after each year’s entry.

At one go, he can fill-in all the years of his career & then make a single Submit.

We must minimise mouse-clicks, otherwise candidates will run-away!

SALARY DESIGN LEVEL PROFILES

Kartavya

Salary Progression Profile / Career Progression Profile

11-08-03

In my yesterday’s note, I said we will make these profiles available both to jobseekers & to corporate subscribers.

To jobseekers:

FREE (as a “carrot/incentive” for them to come online & fill up those content forms).

To corporate subscribers:

PAID (Rs. 10 for viewing / Rs. 50 for printing).
(—after all, we have to make money somewhere—& corporates are the only people who will pay!).

But, we must NOT make available the whole range of profiles (1.02 billion combinations) to the jobseekers.

We are giving away these profiles “FREE” to the jobseekers and therefore, must restrict this “give-away” only to the barest minimum—just enough for him to become “incentivised” to log in & fill up those forms. No more!

And therefore, I suggest that to any jobseeker we make available (free) only following profiles:

Professionals of same Age-Group as himself
Professionals of same Edu. Quali.
Professionals of same Experience
Professionals of same City
Professionals of same Industry
Professionals of same Function
Professionals of same Designation Level

In any case, a jobseeker would want to compare himself (—and, especially his salary & his designation) with his PEERS only.

For a “Manager” to compare himself with other “General Managers” is meaningful.

Or, for a “Pharma” industry professional to compare himself with a “Telecom” industry professional is equally meaningless.

And since, for every jobseeker, we have captured the above-mentioned 7 fields, we can ensure (thru appropriate logic) that he can see/view only these 7 comparative-graphs and nothing else.

Of course, software should automatically add 1 year to his age on each birthday. Same with “Exp” (—although there is a possibility that he might be “unemployed” for last one year!).

Now, of course, there is a possibility (at least, when we upload 100,000 email resumes on our own website), that a Candidate will come back from time to time to “EDIT” his own structured database online.

Q. Should we allow/permit him to simply — and directly — change/alter the structured extracted database WITHOUT also pasting corresponding “revised” email resume?

(Diagram drawn showing four parts: A = Structured Database, B = Email Resume, C = Graphs, D = Keywords.)

If he only makes changes in A without making corresponding changes in B, there can be/will be serious ANOMALIES! He may do this unintentionally or (worse) intentionally! — just to be able to see/view different salary-profiles!

Example:

Suppose, I continue to be a “Manager” but in Part A, I come & edit my designation to “General Manager” (or my “Industry” / or my “Function”).

So that now, I can see/view those salary graphs.

I try to “cheat” the system!

Damage is not so much that the candidate gets to see/view a few more graphs.

The real damage comes when a client shortlists him (—really, software shortlists him) wrongly — based on such “doctored” fields! Entire credibility of our search-engine will be at stake!

So, what do we do?

Can we/shall we say that A / C / D cannot be “edited” at all!

The only way a candidate can effect any change in A, C, D is by changing B! He has to post/paste his revised email resume, if he wants to edit his profile.

Then, in the background, we can ask the software to generate following template:

Comparison of Profile of A.J. Mehta (A):

Field	Based on original email resume dt □	Based on revised email resume posted dt □
Name
Address
Current Co.
Total Exp.	15	30
Designation	Officer	Vice-President
Industry	Pharma	Telecom
Function	Sales	Mfg.
Edu.	B.Sc.	Ph.D.
D.O.B.	9-11-45	11-9-37

Software

Now, if it finds ANOMALIES (we will need to define anomaly for each field) of the kind shown on previous pages, then an “ALERT” message should flash! — and maybe an automatic email will go out to that candidate, requesting an “explanation”. If the “explanation” is not satisfactory, the resume should be deleted from database (of course we must not allot / re-allot his PEN to anyone else. Only an asterisk will show that the PEN has been “blacklisted”).

On our website, even today we have a feature whereby every candidate gets an automatic email (once-in-6-months?) to come & edit his “Resume Form”. This will need to change (the message).

Of course, we also have an automatic email birthday greeting going out on each candidate’s birthday.

A related problem is

— when we convert a candidate’s email resume for the FIRST TIME and find some fields “missing”, how do we get him to fill in the correct “values”?

(1) Can he do it “online”? If so, how?

(2) Can he fill-in “offline” & return “smartmail”?

(Signed off with initials + date: 11/08/03)

Kartavya – RESUMINE

Salary-Progression Profile
Career-Progression Profile (Designation Profile)

At present, we have not made any provision to capture:

Salary
Designation-Level
in Resumine.

We do capture “Current Designation” – but not “designation-levels” over entire career.

Unless we succeed in capturing:

→ “Annual Gross Salary” for each year of career

→ “Designation-Level”

We cannot plot the “Progression-Graph” of a given job-seeker, nor can we compare his graph (growth) with the average/median (Line of Best Fit) for the entire population.

And of course, we cannot plot/compare his graph with following sub-populations:

Persons of same “Age-Group” (20 levels)
“Edu. Level” (8?)
“Industry” (26)
“Function” (53)
“Gender” (2)
“City” (100?)
“Exp. (Yrs)” (20 levels)
“Desig. Level” (9?)

But, if we succeed in:

→ getting candidates to willingly provide such data (preferably ONLINE),

then, we would have hit the JACKPOT!

In such an event, we could have corporate subscribers furiously clicking away to generate thousands of permutations/combinations (approx. 1.3 billion!).

And we could collect Rs. 10/- with each click!

And, we would have spent very little to create this powerful CONTENT (which gets created by the jobseekers themselves).

We would spend very little to maintain/keep this content UP-TO-DATE!

So, we get jobseekers to create a CONTENT which is extremely valuable to a HR/Recruitment Manager, for:

→ Figuring out what Salary/Designation to offer to a candidate being interviewed, without upsetting internal apple-cart (of existing employees).

→ Figuring out what “Annual Increment/Reward” to hand out to existing employees.

→ Restructuring his company’s COMPENSATION-STRUCTURE from time to time to remain competitive in recruitment market.

…line can be substituted with the ACTUAL line (based on 10,000 records).

No one would even notice the difference – and even if someone does, we can simply say that the “AVERAGE” line undergoes constant/continuous change as more & more no. of executives submit their data – which is really so!

To be on the safe side, initially, we may not even show the counter of Population [ ].

We may start showing this counter once the no. crosses 10,000.

Next argument (to convince reluctant candidates):

“With these graphs, NOW FOR THE FIRST TIME, you know, whether you are

Underpaid
Under-designated

Vis-a-vis other executives in our database.

You can use this newly acquired knowledge to:

negotiate with your prospective employer for a better offer
negotiate with your existing employer for a better annual increment (raise).

So, now, for the first time, you are “equipped” to bargain/negotiate from a position of strength/the position of knowledge of a huge, authentic database!

THIRD Argument

Not only you will be able to compare yourself with the entire population of jobseekers in our online database, you would be given “access-rights” to compare where you stand with respect to co-professionals, who are

Of same “Age-Group” as you
Of same “Edu-Level”
Of same “Exp (yrs)”
Of same “Designation-Level”
Of same “City” (where you live)
Of same “Industry” as you
Of same “Function” etc.

Initially, this “access-right” will be granted for ONE year from the date you fill-up & submit these forms.

When you come back & update your own data for one more year (at the end of each calendar year), your “access-rights” will be renewed for one more year from the date of updation.

Kartavya, I believe with these 3 arguments, we should be able to convince 100,000 email resume-wallas (in our database) to fill-up the forms.

Of course, before we send out emails to them, we would need to:

→ Convert all 100,000 thru Resumine

→ Give PEN (User ID) / Password to all

→ Upload on our website, so that when any one of the 100,000 logs in with his PEN (User ID) / Password, he is able to see:

• His own Resumine profile

• The 2 forms which he is supposed to fill in

Pl. feel free to discuss if you have any questions/doubts.

[Signature]

10/03/03

P.S.

Of course, we would need to experiment with the use of the forms & generation of graphs, by creating 15/20 fictitious forms.

12-03-03

Salary / Career Progression Profiler

Along with my note dt. 10/3 I had enclosed 2 forms, for capturing (a) Salary data (b) Designation data.

Whereas the form for “Desig-Level” data – seems fairly simple & easy-to-use (simple radio-buttons), the form for capturing “Salary” details would require a lot of accurate typing by jobseekers. No one would have that much patience!

I suggest following simplified form:

Year	Gross Annual Salary (Cost-To-Company)
	Rs. Lakhs/year (Rs. 00,000)
1995 etc	Two digits max [9][9] • Decimal [9] • Single digit [9]

Assumptions

▶ One decimal point is good enough accuracy

▶ We don’t expect candidates getting Rs. 100+ lakh/year to send in resumes!

With only 3 digits to be typed, it is quite easy.

We could even consider 2 separate “Dropdowns,” one each for:

→ Integers (having 0 to 99)

→ Decimals (0 to 9)

We may not ask candidate to click Submit after each year’s entry. At one go, he can fill-in all the years of his career & then make a single Submit.

We must minimize mouse-clicks, otherwise candidates will run away!

COMPENSATION ANALYSIS

Cyril

Here is an opportunity for 3P to earn big bucks !

And, if we can earn big bucks, then you too can earn at least small bucks !!

Potential is very much there. Question is "How FAST can we exploit this potential ? "

This question is very pertinent Since all databases are "PERISHABLE" Commodities

I am talking about, "Gross Annual Compensation" earned by corporate Executives during 1997- 1998 After many months of struggling to get Annual Reports - and then an equally long struggle to create a database (thru data - entry), we finally have a database of

. Approx 31000 executives

. From 902 companies

The database "fields" are

Name of Company

Name of Industry : (Mostly blank but can be quickly derived from KOMPASS database already created)

Name of Executives

Designation

Function : Totally blank for all records

Co's Product : Totally blank for all records

Annual Salary : (Rs. L/Yr)

Edu. Qualification :

Experience (yrs) :

Age (yrs) :

Commencement of Employment (Date) :

Last Employer :

Last Designation :

I have those which can be used by us in our ANALYSIS

The IDEAL analysis would be graphical - although, in some cases (typical example : EDU, QUALIFICATION), it is simply not possible to draw a graph on the small screen of a computer, when Axis has 50 intervals !

I feel anything more than 10 intervals may be difficult to draw - and impossible to "read"

In which case, we may have no choice but to present the analysis in a "TABULAR" form

In the enclosed sheets, I have shown how graphs & tables may appears on the screen.

Although Sajida has started preparing some tabulations, after looking at

. The number of "Population"

. No. of variables in each "Population" and therefore .

. The no. of "Permutations/ Combinations" possible.

I have come to the conclusion that it is IMPOSSIBLE to prepare hundreds of thousands of tabulations manually.

. Leave along drawing "graphs" for each of these tabulations !

. So, presenting our findings in the form of thousands of static HTML pages is out of question.

. What we need is a SOFTWARE that would "dynamically" generated each of these tabulation/ graph depending upon the QUERY, and display the same.

Now,

I am NOT talking of having this software/ functionality on our WEBSITE.

Then we don't make any money !

I am proposing to put all of these (database + Software) on a CD (of course, securely encrypted) and selling each CD to corporate clients for say Rs, 30,0000 / = each.

If we can find just 100 buyers ,this means of Business of Rs. 30 Lakhs.

To market this product, we would need to

. Advertise heavily in Newspapers/ Magazines

. Send out 10,000 + mailers

. Follow - up over phone

. Give "demonstration" seminars to Personnel Managers / CEOs in 5 star hotels (followed by drinks/ dinner)

But if we come - out with a good product and come out QUICK - then there is a good chance of Success. Of course, we would need to RISK a few lakhs of rupees on "Promotion" but then risk is an integral part of any "PIONEERING" Product.

Since the database pertains to 97-98, it is already ONE year Old ! 98-99. Annual Reports would start coming out in another 2/3 months. Thereafter, "Customers" might lose interest in such old data. Hence we must come out with this products within 2/3 weeks.

Perhaps it may not even be necessary to write a "custom" software. Could we possibly integrate some "standard statistical software package such as

. SPSS

. STATSOFT

I believe these packages cost between $ 500 - 1000.

Will we need an API ? And solid ENCRYPTION is a MUST although, we must provide facility whereby a Personnel Manager/ CEO can take a hard copy "PRINT - OUT" of any particular "Graph/ Tabulation" & be able to circulate amongst his colleagues for comments/ discussion

Of course, no one would have time/ energy to print - out thousands of combinations.

I would like to discuss this with you URGENTLY.

h.c.parekh