The data migration market is there and growing

But it is not in the skull of the decision makers

The proof: 10 years ago 80% of the migration projects were unsuccessful. In 2007 after the survey of Bloor Research, 84% of the migration projects failed: either overrun the budget or could not finished in time. The survey also says that the budget of the data migration projects will raise from 500 to $ 900 million until 2012.

On the 1st October I participated on the event Data Migration Matters 2 in London. In the next posts I reflect my thoughts about the lessons of this day.

Why migration projects fail?

Migration is always the necessary evil in introduction of a brand new shiny system or organisation restructuring. That is the reason, that they are always underestimated, and the scope is not completely defined.

To save costs, companies think that the cheapest solution is the “DIY” – Do It Yourself. The result: 89% of the internally made migration projects will fail. I do not mention now the telecommunication companies, banks and insurance companies, where data migration projects are continuously executed, I  mean the companies, where a big restructuring or introduction of a new ERP happens once in 10-15 years.

If decision makers think about the migration of existing data, than they give the task to the IT department. It don’t want to underestimate and decrease the significant of the IT in a company, because I have also strong IT roots, but this time I must clearly say: data migration is not the task of the IT Experts! The thinking of the IT people do not reflects the Business Needs and the result will be always unhappy business users and/or failed project. The solutions is building virtual teams from Business Experts, IT Experts and Project Manager. Johny Morris, the author of the Bible of the Data Migration (Practical Data Migration) thinks: the data migration project is a quest, where all parties have a clear goal to finish the project successfully. In a quest all team members will give their best to achieve the common goal.

The last significant failure cause of the data migration projects is the lack of specialists. There are lot of independent consultants (like myself on migration.hu), migration consultancy companies (like iergo, Kognitio, Datamonic, DataFlux), solution providers (like Talend, Trillium Software, Informatica) on the market, they have the knowledge and proof of finished successful migration projects.

What is the goal of the migration projects?

Every project manager want to hold the number of risks low in their projects. At the moment an application migration project has a very big risk: the data migration subproject. The goal is to lower the risks of these projects, to predict the outcomes of the data migration, lower the impacts in a daily business activities. These all will result a Zero Defect Migration which is the main goal of a data migration project.

Conclusion: Supply and Demand

Why do not meet the highly educated and experienced Migration Solution Supply  and the Demand in the market? The answer is the bad timing! In 98% of the life cycle of the companies has no need to migrate data. And when they need to execute data migrations, than they don’t know about the safest and cheapest solution: using of data migration experts.

So our task is in the next years to communicate to the world: there is a solution, there are proven ways, better to estimate first an unknown task in case of data migration needs, than fail and loose lot of money.

There is a solution to eliminate the bad timing of the migration in the market and lower the failure risk in a potential data migration project: it is already being used: keeping the data quality of the data centers in the companies by using data governance.

Once again: the data migration market is there. We, the members of the data  integration profession community can deepen the market in the head of the decision makers with continuously communicating with the potential customers.

Bookmark It

Hide Sites

How becomes a data migration professional to a victim of a failed migration project?

The way NOT to handle your customers

I changed the role from from Inspector to Robber this weekend: one of my webhosting provider announced a server change on Saturday which might have cause 2-3 hours outage.

I did not care too much about it, I have only some test domains stored there. However after the promised maintenance time it began to be suspicious for me when I realized something is stinky here.

I realized three mistakes, which can not particularly strengthen the customer – service provider relationship:

• service outage
• plan only for nice-weather fall
• bad communication

Customer service mistake Nr. 1

I wondering who has been asked whether the Saturday is acceptable for a 2-3 hours total outage.

Imagine the following hypothetical situation: You are the webhosting provider of Facebook (300 million user), and you send an email for your only customer: “Dear Customer, on the 18th of September we will change the transformer relay for our servers, and it may cause 3 hours outage, because of possible electrical power outage. “.  What do you think, how long will it take to change the provider?

When I have been worked at Siemens, 10-20 people have used a test system from more different countries. If one of us wanted to execute an action which had an effect to the entire system, then it was a must to send out a message and wait for answers, whether everyone accepts the system outage.

Webhosting service can be financial very sensitive for the business of the customer – so a total outage for many hours is totally unacceptable. If there is no idea how to avoid the outage, the I would recommend to do the maintenance in the night hours, where a visitor activity is more unlikely.

Customer service mistake Nr. 2

In the evening of Saturday I could not receive emails from that particular domain (it was after the announced finish time of the maintenance), I got always the message from the e-mail client, that the login authentication was not successful.  After the third-fourth attempt I checked my domain: it was inaccessible  neither from the old IP-Address nor from the new give IP-Address. Nice, I thought, but I was patient and my e-mails are not particularly important on the Saturday evening, so I waited until the next forenoon.

You can find out: yes, the service was still non-operational, so I have written a short mail to my provider. They answered very quickly – a positive sign.

But after 1 day the they had no idea what was working on the system and what not, based on the question to my claim (that nothing works).

I have to admit, the service was re-established in 1-2 hours, and everything has worked again.

The mistake here: the provider had only reacted on the complaints of the customers, they had unexpected errors which means they had no plan B for the changed situation. However the guys could solve the problem in conceivable time with improvising.

Customer service mistake Nr. 3

Communication. I received a message from the provider that the migration was successful.

Well, I don’t call it successful when the outage took 1 day instead of the (also unacceptable) 2-3 hours. And what is worse: They thank those who had helped them and not “molesting” them during their work.

That means there were some they dared to ask what is going on with the website.

This  unacceptable arrogant voice – sorry – is not the way to keep the customers and even get new one by reference. It should have been avoided, such sentences are absolute unnecessary in a customer – provider relationship.

Conclusion

I am wondering, whether the provider had a though about these questions:

• who guarantees, that such outage won’t happen in 1 year?
• who guarantees, that no important data has been lost?
• who guarantees, that no important e-mail has been lost?
• who takes responsibility about the eventually lost revenue, If somebody runs a e-commerce on the page?

I was thinking to buy a new storage in the near future, but because of the communication to the customers, I will choose definitely another service provider.

In the next part I will write about some solutions, where the outage could be avoided by

• existing commercial solution
• or find a smoother way to reduce the outage risk

Bookmark It

Hide Sites

How to succeed a (2-day) data migration project?

Contact list data migration part 2

You can laugh about the volume of the project, but if you cannot succeed a short project, how do you think you will be successful in a large project?

This post is the second and last writing of the  The true story of contact list data migration, where many contacts from 4 different sources are migrated in a possibly  maximum  data quality to iPhone.

Transient data store for migration Nr.3

You saw in the last post, that I had problem in the “last name“-”first name” sequence in Google Address List, and there is no way to repair it automatically (which in the end effect faster then manual correction). The next attempt is to export all the contacts into a CSV file, which can be imported into Excel (or in my case OpenOffice Calc), where the name sequences can be repaired relative quickly. This is already the third transient data store.

Note: At this point I can check off 2 legacy sources: contacts from the old Nokia phone and from the Thunderbird contact list. The Excel table with old numbers and an Outlook Express Address Book is remaining.

An obvious choice is copy-paste the Excel contacts to the OpenOffice Calc (in CSV format), of course with merging the right columns.

The contacts in Outlook Express are in the WAB (Windows Address Book) – which can be imported into a CSV format – so I have again a direct path to the end list with copy-paste.

The last steps as importing back the Google-kind CSV format to the Google Contact list and final synchronisation with iPhone through the iTunes is without any problem and I can be happy.

The key element of the migration

After three attempts I have found the safest and quickest method (and transient data store), where the data import, data modification and data quality optimization has the fastest way: using the Google Address Book CSV format in the application OpenOffice Calc. Look at the final workflow which became complicated in the first look.

Data Quality Aspects

The original scope in aspect of Data Quality is the maximum data quality, there are no excuses, I don’t want to use any garbage in the new phone. Here is the hypothetical question: what is maximum data quality? All data records in the target are fine, so I can use them? Or: all data records in the source(s) are in usable form in the target? Of course not, there are many useless e-mail addresses in the sources which have been added automatically after writing a mail – even from 5-10 years ago, where I don’t know any more, who is behind the e-mail address.

Here are the important points in the data quality:

- the sequence “last name”-”first name” which was explained very detailed in the last post

- the using of the Hungarian special characters as: ö, ő, ü, ű, í, á, é, ó – fortunately using the UTF8 coding between the export-imports that was no problem.
- The right source fields should be placed into the right target fields (e.g. it has no sense a phone number in the e-mail column)

- The phone number formats should have this pattern: (+<international_code> <national_code> <phone_number> )

Elements from Project Management

If I were enough schizophrenic, then I could have the following concerns for a meeting:

- The rough effort estimation said: 2 days time for this activity.

- The risk was mentioned in the beginning of the first post: the biggest risk is not to reach the project goal within time, because of unknown functionality of the available tools. And this was the main focus during the 2 days: to find the fastest solution to merge the contacts. The risk evolved to a real problem, but it has been handled by focusing the solution.

- Measuring problem: How could I be certain, that all important contacts have been migrated? The usual counting technique (counting of the contact of each source and comparison with the number of the final migrated contacts)  is not efficient because of many useless e-mail addresses (with the pattern info@xyz.com). What do you think, what would be an acceptable method to be sure not to forget any contact?

Summary

I declare this mini-project successful, because the migration has been executed in time and in the target point of view with max. data quality. However there were lions in the path even in this relative simple and short activity which endanger the reaching the main goal.

Bookmark It

Hide Sites

Project contact list migration part 1

My old Nokia phone reached this year the school-age (became 6 years old), and I was satisfied with him in the past. Unfortunately it became deaf and mute recently, so I was forced to purchase a new phone (iPhone).

I possess a mobile phone since 1997 and I carried the contacts from phone to phone. Beside this I have an Address Book of the E-Mail software (Thunderbird) with contacts, an address book of Outlook Express from an old PC, and an Excel table which is a copy of my former business phone.

So I have 4 sources with various contacts (redundancy is there), and a question: How to merge and migrate all contacts within the shortest time and maximum quality? (in this question I have to be maximalist: I do not accept 95% data quality, because what should I do with 20-50 garbage contact in my new phone?)

Migration Planning and Project Initiation

After the scope is defined very clearly, I form the topic of effort estimation generously this time: max. 2 days.

Configuration Management and software tools:

Because typing on the PC is 10 times faster than with the tiny buttons of iPhone, I have to find 1 or 2 transient data store, where the synchronization can happen very fast. For the last step iTunes synchronisation tool is the only appropriate way and it works properly.

Plan of workflow

The key element of the workflow is the right choice of the transient data store. Because the export-import possibilities of the different data sources are unknown at the moment, this key momentum will be clarified during the project.

Because of this we found in this tiny project a risk which can affect the delivery time (in this case the time of full contact list on the iPhone) – and as we know the most of the data migration projects (84%) fail because of delayed delivery time.

Transient data store for migration Nr.1

The extraction of the contacts from the Nokia is theoretically very simple: I have no tool, no data cable to download the contact list – so the manual job remains whatever it hurts. The destination: plain paper or an address book directly. I have chosen the direct entry into the Thunderbird Address Book, because nowadays I type faster than write. And with this move the number of data sources will decrease with one.

Surprise Nr 1: There is no direct connection between iTunes and Thunderbird. The possible contact sources of iTunes are Windows Address Book, Google Contacts and Outlook Express Contacts.

So lets export the data from Thunderbird to Windows Address Book. Of course there is no direct connection in this direction, so I have 2 possibilities: CSV export and LDIF (LDAP server).Note that at this time we have 3 transient data sources: Thunderbird Adress Book, CSV (or LDIF) and WAB.

Short summarized: about 30% of the contacts were simply not migrated, I don’t know the reason. After one or two attempts I dropped this possibility and have chosen the remaining option: Google Contacts

Transient data store for migration Nr 2.

The life is not easy, there is no direct connection between Thunderbird and Google Contacts. After a short research I found  the Zindus add-on for Thunderbird, where the synchronisation between Thunderbird and Google Contacts supposed to be solved.

And it worked: the data synchronisation between Thunderbird and Google and also between Google Contacts and iPhone through iTunes.

Surprise Nr 2: However as I checked the application contacts in iPhone, I faced a new problem: the sequence of last name and first name.

For explanation: Hungary is one of the rare places of the world (beside Japan I think), where the official sequence of the name is “Last name” “first name“. So I want to see my contacts in the phone that way, however Google Contacts messed this sequence up.

Of course the Google application does not recognise (even if the settings are for Hungary), what is the right sequence. The fast solution is put a comma between “Last name” and “first name“, like Bossányi Tibor. And this is again manual work for many hundred contacts. (I don’t want to tell long stories about the attempts to set the settings to English (UK),exporting the list in CSV to find the fastest solution to put this comma into the names. It does not work.)

Summary

To not to loose too much time with this simple operation (at the first look), I had to think in a project way: finding a clear scope, a simple effort estimation and a  measurable action plan to avoid shifting this activity. In this first part I found the most optimal way between the possibilities by attempting more different ways and measure which is the most effective in aspect of time and data quality.

The next (last) part I will write about the data quality rules which have been set up, and the migration from remaining legacy sources.

Sources

Data Migration Project Checklist from datamigrationpro.com

Bookmark It

Hide Sites

,

What do you do with the bounced e-mails after sending some thousands to your customers?

Every enterprise which stores a mass of data meets the Loch Ness-monster of the data handling which appears from the nothing and you cannot catch it. If the newsletter auto-responder system could cry (or even swear like a trooper), then you have to close your ears which is it worth nothing, because the noise comes through the ear-plug: “this is not a database, this is a dump“.

Why is a clean database important? Because the online marketer cares about the conversion, measures the percentage of the number of buys / sent e-mail and it matters if only 0.1% or 30% does not find the target address and fail to land in the right e-mail account.

Why does so many error arise in the customer database?

Although the (good) online marketer try to solve the subscription form as simple as possible (name and e-mail address), there are many reasons why will be useless the data at the end.

- the customer puts a wrong e-mail address into the form

- a bad name is given – it is important, because in most of the cases the marketing e-mails are personalized

- The provider of the target e-mail address judges the sender address as a spam

- if the customer reaches the level of buying attend, and he gives an erroneous the shipping or billing address

- e-mails are bounding back because of full mail box.

Explanation of the colors:

blue: a better newsletter and auto-responder software can handle these errors – they are syntactical errors

green: to eliminate these errors the owner of the data is also needed, who knows what should be corrected – so an automatism is excluded (they are semantic or content failures).

What can the online marketer do to enhance the data quality?

1. Nothing: It is also a solution, but of course a lot of money can be lost
2. He maintains the database regularly: The maintenance can be done manually, but we all know that most of the online marketers and even their webmaster aren’t bit artists so they can spend 1-2 days to cleansing of the data. A better solution is using a software special for this area, with this the task can be completed within some hours – after the determination of the correction from error patterns. However I must highlight again, do not imagine the whole job so that the customer list is fed in the input and a perfect list will be spit out in 2 hours! It does not work that way. In many cases the decision of the owner of list is needed how to correct some sort of data error.

If you have a question about the maintenance of the customer database then please write a comment or contact me directly through the “contact” menu.

Bookmark It

Hide Sites

I mentioned the 3-dimensional layout of the data profiling in the last post (Data Profiling).

The column level has been depicted quite detailed. Because the description and the practical example the second dimension caused some questions, I devote a deeper analysis of the intra-table (table-intern) level using the open source Data Profiler of Talend.

Normalization

The goal of the data profiling is not only to detect the user-created errors, but you can also determine the logical design defects, which can cause duplications, redundancy in your database. The intra-table (used as table-intern in the last post) profiling level is related with the dependency checks within the table. With this the existence is of the 2nd normal form of the table design is checked, in other word if there any columns in the table, they are dependent from a non-key column.

This table is not fully normalized for the second normal form, because the columns City and Country could be stored in an additional table and they can be saved as foreign keys in the table Address. However it was my decision not to create to many tables in the database, and the creation of the full addressing is also easier and quicker. However it does not fit to the principles of the second normal form. So the column City is dependent from the non-key column Country and it could contain defected entries by the user.

Theoretically the column PostCode is also not normalized for the 2nd normal form, because it contains not unfinite combinations and it is also depending of the column City. In the next example we will check this column by creating the Data Quality Rule in the Talend.

Data Quality Rule for the PostCode

The definition is for this rule is the following: ‘All PostCodes from the City Budapest are beginning with 1 and have the length of 4‘.

For this reason the Where klauzel of the DQ-Rule is: City = ‘Budapest’ and (left(postcode,1) != ‘1′ or length(postcode) != 4)

The quality of the table in aspect of this rule is 99.76%, that means we have a few defected rows, they do not match the criterium.

In the last step let’s check the defected rows, where we can see, there are NULL values for the column PostCode, there are non PostCode-like entries and there are some which begin with a whitespace or end with a dot.

Summary

The creation and execution of this Data Quality Rule took me 10 minutes. However you can see, for a large database you can define hundreds and tousands of  these rules which can create some man-month work. The bigger effort is the resolution of the data quality issues – where you, as the leader the data migration project, cannot decide alone, which errors are important and which not. But the detecting of such errors is worth it to sacrifice the effort.

Bookmark It

Hide Sites

You were probably in the situation when your furniture had to be moved from their old place because of relocation to another city or before flat painting. You move the floor-rooted piece of  furniture and you realize that the back is covered with discusting spiderweb and other grime. What will you do: leave it as it is, because it cannot be seen near the wall or you will remove it immadietly?

Next time you will reckon and check the corners, not visible places of the room what should be removed what does not match into the room. That means you execute a dirt profiling from time to time in your flat.

A similar approach must done during the data migration project. Checking the data quality is one of the first steps in the project: with the execution of the data profiling can we build a first impression about the scale of the project. That means this step must be executed before the final effort estimation!

What is the data profiling?

During the data profiling process you will examinate the data in the legacy source to collect statistics and information for building the data quality rules. For the dimensioning of the definition of the data profiling, I have chosen the determination by Informatica: The data profiling has 3 dimensions:

- column level

- table-intern level

- inter-table level

Metadata profiling

The first two levels (column level and table-intern level) can be examined with the metadata profiling. On the column level you can check the

- data types

- domain, range of the values (i.e. post code must be within the interval of 1001-9999 in  Hungary)

- pattern (i.e. the phone number has the pattern: +nnWnnWnnnnnn)

- frequency counts (i.e. most of the sells happen on workdays: Tue, Wed, Thu)

- Statistic numbers (min, max, median value, avarage value, etc)

- dependencies

- Redundancy

You can draw information within the table by dependeny checks (this category takes also place in the third dimension: in the inter-table level). You can determinate the dependency between column values by the normalization rules from the logical data model design: i.e. a national code of a phone number is related to the ‘city‘ column.

Finding dependencies between the tables (inter-table level dimension) are based on table model design: i.e. a foreign-key value customer-id in the orders table must appear as primary key in the customer table. After my experience the most the data garbage is coming from the missing referential integrity between the tables which was caused by poor data model design.

Example by an open source tool

I have chosen the Data Profiling Tool by Talend on a database with some tousand records of the ‘Address‘ table. The free downloadable version supports the first two dimension of the data profiling types: column level and table-intern level.

In the picture below you see the example of two columns: Address and AddressID. The meaning of the colors of the column Address.

• Red: number of all records in the table (6575)
• Yellow: NULL values
• Orange: Distinct Count (6180)
• Blue: Uniqe Count (5910)
• Pink: Duplicated Count (270)
• Light Blue: Blank Count (39)

The Primary Key AddressID seems to be OK, because the count of the Unique values are the same with the count of the rows. But what can we do with the column Address? Theoretically it can be also uniqe, however the column contains street and number.

You can see in the picture above the defect rows. There are clearly bad administrated data, instead of the real address we find city name, phone number and duplicated addresses. To eliminate the duplicated rows the connected tables of address must be also checked by the data profiling tool.

In this short example you could see more data profiling types: redundancy for the duplicated rows, range for the recognized phone numbers, frequency counts, etc.

Summary

The data profiling is the anteroom for the creation of the data quality rules. To execute the whole data profiling, you have to check each column, each table and each connections between the tables. If you have the statistics and all information about the defects, then the data migration team, where the right stakeholders from the business side are also member of the team, must decide about the measurements and the invested efforts for the fixing.

References

Informatica

DataQualityPro articles

Wikipedia

Bookmark It

Hide Sites

Data Migration Market Research

Why did I shoot full the Hungarian Internet and video sharing sites in the last months with the message, that the Data Migration is an area which is not considered by the industry analyzers in the most of the cases?

Lets look the boring numbers: The Data Migration part of the projects has a success rate of only 16%, in other words those, they are delivered in time and in the budget limit*. (*The facts and numbers are based the Bloor Research from 2007, author is Philip Howard Research Director)

Stop for a moment and look what does it mean. Imagine, that your boss gives you 10 tasks in 10 months. You will accomplish the task in 10 months once or max. twice, that you don’t run out of the time and you don’t spend more than planned. What do you mean, is your job ensured after this efficiency? Not to mention, if you have a company and 16% successful closed projects…

Don’t take this as bagatelle, because data migration should be a niche market.

1. After my research Data Migration is necessary at least in 50% of the application development projects.

2. The Data Migration is the part of the bigger project in most of the cases, and the number above is the bottle neck. So if the task data migration is not successful, then it will pull down the entire project.

3. After the Bloor Research’s survey the data migration projects reached the threshold of 5 billion $ in 2007. The market is growing and until 2012 it will reach the $ 8bn. This is a growing of 12% yearly in average.

What is the reason of the high rate of unsuccessful data migration projects?

If you take your 15 years old car to the garage for yearly checking, the experienced mechanic will say after the first look , that your car will be ready in two days, and he will specify what will be checked and change. What will happen if you call your mechanic on the third day?

“Mr. Bossanyi, this car has much more problems than expected. I can clearly see, that it suffers from missed maintenance. Furthermore there are some defects that I cannot repair on-site because I don’t have the necessary instrument, I have to take it into the “Super Diagnostic Garage” to find the reason to get fixed. And the worse thing is, it has some electronic problem which is far not my area. I could not find any expert who help to solve this problem. You have to take your car in some “Hyper Diagnostic Garage”.

How to translate it to the world data migration?

1. The task was estimated the badly, the real job is much more than is seemed in the first look
2. The scope determination was insufficient
3. The necessary tools and instruments were missing
4. The responsibility area was not clearly defined
5. The communication between the experts / stakeholders was partly missing
6. There are missing actions from the past which have been ensured the high and sufficient quality of the system (in the example above the car maintenance).

These points are the most significant problems, they cause the miserable results of the most data migration projects. The goal of this blog is to share my experiences from this area and find solutions to reach a better success rate for the next market survey in 2017. If you would like to find more information from the data migration profession then visit datamigrationpro.com. If you want to know more from the related area, the ensuring of data quality issues, then visit the community site dataqualitypro.com.

Bookmark It

Hide Sites

Mr. Wolf is a Problem Solver in the movie Pulp Fiction. His task  was to solve the big mess left after a gun fired accidentally in a  car.

The Data Migration Expert is also a problem solver. His task is to solve the big mess between application development, customer requirements, verification – integration, data owners and domain experts, to find an optimal quality of data which will be loaded into the target system, and to find an optimal concept how to achieve the cut-over.

At the present time am looking for data migration projects in the European market. I see lot of requirements what the right candidate must own in his experience history.

What importance is attached for a data migration project leader ?

In the most of the cases you should know the target system, and you must have more years experience. Surprisingly the source system is rarely a requirement. In addition

• the industry environment (insurance, healthcare, etc.)
• the target database
• in most cases he must know the programming language of the application or migration tool.

These are all as essential requirements to get into a data migration project.

You have (only) an advantage if you know the Data Migration Methodology and have experience in a previous project.

These requirements are very technical-oriented, and they did not show to a successful migration delivery. If the requirements for data migration project leader are focusing only to the technical details, then the project can eventually end with a rescue mission.

What should be important for a data migration project candidate?

It is a good idea to know the target system and the business cases of the target system, as mentioned in the most of the project announcements.

Knowing the business cases not equal knowing the workflows within the enterprise using the particular business case. To find an optimal way to execute a smooth migration you have to be clear about what the processes in the enterprise are and how do they connect with each other. If you see the single business cases, you won’t understand the whole picture which can lead to mistakes.

My advice for the leader of a migration project: look at the big picture, see the whole system, dive deep into the processes.

My advice to the recruiters: knowing the programming language of the migration tool is not enough, even knowing the the functionality of the target database is not enough. To know the target system, how does it works, what are the use cases, is a little better. An optimal choice is to find a person, who is able to see how to match the new target system into the workflows of the enterprise during the migration steps. Someone, who does not forget to adapt the legacy system intelligence into the intelligence of the new target system. Good luck in finding this skill set!

Bookmark It

Hide Sites

Az adatminőség-pokol harmadik köre: a technológiában tivornyázók köre

Dante Poklának harmadik körébe a tivornyázók, torkosok és falánkok kerültek. Az adatminőség-pokol harmadik körébe azok tartoznak, akik a technológiában dőzsölnek, ettol az eszköztől várják az üdvözítő megoldást.

Azonban nem azok a vállalatok az igazi bűnösök, akik felhasználják a különbözo technológiákat ahhoz, hogy a vállalaton belül az elfogadható minőségű adatokkal támogassák az üzleti folyamatokat. Ok áldozatok.

Habár megfontolandó kérdés, hogy a CIO és CEO (műszaki igazgató és a ügyvezető igazgató) mit érdemel, ha bedől egy ügyes (vagy nem elég felkészült?) értékesítőnek, aki elad a vállalatnak egy csodálatos szoftvert, amelyet

• könnyű installálni
• felhasználóbarát kezelőfelülete van
• a tranzakciókat nanomásodpercek alatt elvégzi
• pillanatok alatt felfedezi és az adatminőségi problémákat
• és nem is engedi, hogy az üzleti folyamatokban adathibák keletkezzenek.

Ez a rendszer önmagában nem elég, nem oldja meg a problémákat. A legfontosabb, hogy a vállalaton belül tevékenykedő emberi lények közreműködjenek.

Emberek

Emberek kellenek ahhoz, hogy megfogalmazódjon a gondolat, hogy a vállalat működésében felfedeztek egy problémát, amikor olyan adatokkal kell dolgozni, amely inkább csak hátráltatja a hatékony munkát. Ezt a problémát meg kell valamilyen módon oldani. Felállítanak egy csoportot, amely az üzleti oldalról az adatok tulajdonosai, használói vannak, az IT oldalról azok, akik a folyamatokat biztosítják.

Le kell győzniük a félelmet, hogy a esetleg blamálódnak a hibák okozása vagy sikertelen kijavítása miatt. Figyelembe kell venni az emberi faktort, mert emberek fognak részt venni projektben, és az ő sikerük fogja meghatározni az egész projekt sikerét, nem a használt technológia maga.

Folyamatok

Az üzleti folyamatokban áramló adatok karakteriszitkája vállalatról vállalatra különbözik. Sot ugyanabban a vállalatban a különbözo funkcionális egységekben is mások lehetnek az adatminőségi követelmények. Más üzleti szabályok érvenyesülnek a különbözo projektekben is. Ezeket tehát figyelembe kell venni, amikor meghatározzák a hatékony módszertant, amely segít sikeresen véghezvinni az adatminőség javításhoz létrehozott projektet. A legjobb módszer maximalizálja a ráfordított idő- és költségkeretet.

Technológia

Végre. Talán megkönnyebbülsz, hogy a hatékony technológia is szükséges a célok eléréséhez. Konkrét megoldásokat most nem említek ebben a bejegyzésben, mert általánosan megfogalmazott problémára nem lehet konkrét, kiragadott megoldást megemlíteni. Létezik open source és nagyvállalatok számára rendkívül komplex és drága technológia is.

Összefoglalva komplex problémákra nem létezik gyors és egyszeru megoldás. Nincsen varázspirula technológiai megoldás (válaszd a kéket vagy pirosat), és nem is lesz. Egy szervezet útkeresése az adatminőségi problémák megoldására csak akkor lesz sikeres, ha egy emberek alkotta motivált csoport a helyes módszertant alkalmazva felhasználja a legjobb technológia nyújtotta lehetoségeket.

Ez a bejegyzés az “Adatminőség-pokol 9 köre” sorozat 4. része volt.

Bookmark It

Hide Sites