This is a common question that I see people asking in the industry. The next question usually they ask - I already have my data in good shape - why should I pay for cleansing it.
My question to all these answers is - YOU DONT KNOW HOW IS YOUR DATA.
Why I say that? Because while I am playing with lot of enterprises data in last 5 years i come to know from interactions with customer or people who are maintaining the data this fact. Especially if there are multiple departments handling all kind of data and they dont have a cohesive mechanism to deal with the data as a single form of truth.
Now whatever be the reason for your company - its indeed true that data is not in a good shape that it should be.
How it effects - you dont have visibility towards your spend properly / by commodities / by supplier groups / by geographies or by negotiating power that you are carrying and last one is risk that you are maintaining.
To get this visbility you need a dashboard which gives you analysis power. To get a dahsboard you need data standardized, normalized and enriched classified so dahsboard shows correct reports.
To get normalization, classification task done you need a process or product or provider who provides these services rather than bulding any of these at your end. An here the debate starts - who, how much and when, what ?
I start my day by looking at lot of web posts, blogs and news where people tell you horrible stories about hard days, saving money tips and spend analysis, innvovations in sourcing, optimization and what not. Everybody is telling you that you need to analyze your spend and how one product is good than other product to do this. Here are my two cents -
First - Its your spend and your sourcing data. So there is no readymade packaged tool or product that understands your data end to end. So whoever is saying its fully automated solution is telling a lie. What you need to buy is a packaged process - whether its a driven by a expert team or driven by a software tool. The process is more important.
How do you decide that one process is better that other process for classification? there are few parameters to decide. Most important is how that process looks at your data. I will try to eplain this more in next 2 days.
You can post your comments questions here or mail me on pmendki(at)gmail(dot)com
- Prashant Mendki
My question to all these answers is - YOU DONT KNOW HOW IS YOUR DATA.
Why I say that? Because while I am playing with lot of enterprises data in last 5 years i come to know from interactions with customer or people who are maintaining the data this fact. Especially if there are multiple departments handling all kind of data and they dont have a cohesive mechanism to deal with the data as a single form of truth.
Now whatever be the reason for your company - its indeed true that data is not in a good shape that it should be.
How it effects - you dont have visibility towards your spend properly / by commodities / by supplier groups / by geographies or by negotiating power that you are carrying and last one is risk that you are maintaining.
To get this visbility you need a dashboard which gives you analysis power. To get a dahsboard you need data standardized, normalized and enriched classified so dahsboard shows correct reports.
To get normalization, classification task done you need a process or product or provider who provides these services rather than bulding any of these at your end. An here the debate starts - who, how much and when, what ?
I start my day by looking at lot of web posts, blogs and news where people tell you horrible stories about hard days, saving money tips and spend analysis, innvovations in sourcing, optimization and what not. Everybody is telling you that you need to analyze your spend and how one product is good than other product to do this. Here are my two cents -
First - Its your spend and your sourcing data. So there is no readymade packaged tool or product that understands your data end to end. So whoever is saying its fully automated solution is telling a lie. What you need to buy is a packaged process - whether its a driven by a expert team or driven by a software tool. The process is more important.
How do you decide that one process is better that other process for classification? there are few parameters to decide. Most important is how that process looks at your data. I will try to eplain this more in next 2 days.
You can post your comments questions here or mail me on pmendki(at)gmail(dot)com
- Prashant Mendki
Comments
Post a Comment