How to Overcome Your Data Quality Superstitions

#DataBadLuck

How to Overcome Your Data Quality Superstitions

Donato DiorioFounder & CEOBroadlook Technologieswww.broadlook.com

Michael FarringtonChief Product OfficerRingLeadwww.ringlead.com

#DataBadLuck

Collaborative selling

CloudBig data & sales

intelligence

Mobile

Key trends in CRM

Social

Analytics

Metrics/ Dashboards

Empowered Users

#DataBadLuck

Enhance

Clean

Protect Enhance

Without…

Limited

potential

Without enhancing your existing data

you limit your “data potential”

Decaying data

Without a protection

strategy, your data will continually

decay

Poorfoundation

Without performing a

comprehensive data cleanse, the

foundation is weak

The Foundation: Good CRM Data

#DataBadLuck

It’s ok to delete data

MYTH

#DataBadLuck

Never delete CRM data - Score it Instead!

[email protected]

[email protected]

#DataBadLuck

Scoring Data: Focus on Good Data

[email protected]

[email protected]

[email protected]

#DataBadLuck

What can we learn/derive from the existing data?

• Domain is broadlook.com• Email pattern in first-initial(.)last-name• April reports to Donato• On broadlook.com, there are 15 additional contacts• Notes on Donato are 1 month old• Notes on April are 8 months old• Natalie is no longer at the company• “The Doctor” is a fictional character• Natalie is now a VP at another company - and an

additional prospect!

#DataBadLuck

Result: A More Complete Picture

[email protected]

[email protected]

[email protected]

[email protected]

[email protected]

[email protected]

[email protected]

New Prospect:

#DataBadLuck

MYTH

Using the stick works:make all fields required!

#DataBadLuck

Using the stick works

• Determine carrot and/or stick on field basis, not object

• Educate users on the importance of everything you ask of them (focus on selfish reasons)

• Don’t ask what they don’t know

#DataBadLuck

Training works to enforce data standards

MYTH

#DataBadLuck

#DataBadLuck

Enforcing Data Standards is Optimal

#DataBadLuck

My Data is Fairly Complete

MYTH

#DataBadLuck

My Data is Fairly Complete

• Superstition or fact? Find out!

#DataBadLuck

My Data is Fairly Complete

#DataBadLuck

Buy as much data as you can, all at once

(because it’s cheaper)

MYTH

#DataBadLuck

Data decay happens

• Change in title, promotion

• Change in working location

• Change of phone number

• Add mobile phone number

• Change of department

• Change of area code

• Change of email format

• Merger or acquisition

#DataBadLuck

#DataBadLuck

Bad Data is IT’s Problem

MYTH

#DataBadLuck

Bad Data is IT’s Problem

• He who reporteth upon it...

• Treat it like a project

• Choose Data Quality applications that don’t require a PhD in Computer Physiology

#DataBadLuck

The company’s name is more important than

the website address

MYTH

#DataBadLuck

Contact based

Month

NamesTitles

Emails addressPhone

BiographiesSocial Network Links

Real time content

spidering

Event & Activity Based

Day Hour

Dynamic

NewsEmail content

BlogsNet links

social networksnewsgroups

TweetCheck-In’sProximity

Website visitsEmail reads

Semantic monitoring

services

Real time API’s

Company Based

Decade Multi year Year Quarter

Static

URL Corp Name

CityState

AddressZip

PhoneCompetitors

RevenueEmployeesProductsServices

Financials

Database merging + algorithm

Editorial & Aggregation

Editorial + SEC spidering

Changes

Data types

Acquisition

method

Static, compiled and online databases Real timeUpdate

strategy

#DataBadLuck

I know how to search for duplicates

MYTH

#DataBadLuck

I know how to search for duplicates

• It gets messy

• Users may not have access

• Even if you do, is that a good use of your

(user’s) time?

#DataBadLuck

My vendor’s data is better than mine(they are the specialists right?)

MYTH

#DataBadLuck

•Buy data from multiple sources•Refresh top companies with editors •6 month cycle (top 10K companies)•6-12 month (next 40K companies)•24 month cycle on the next 2 million•Nothing past the top 2 million•Add social data (good for top 10%)•Add news feeds (good for top 5%)•Mob source

Data industry processes

#DataBadLuck

How recent is the list as whole? How quickly was the list produced? Different from record freshness. Contact data degrades 3% per month (5% in a stressed economy). A list of 1000 records can be built over 60 days. In the case below, the first 500 records are 8 weeks old (5.68% inaccurate) upon list delivery.

96.8%

Buying data...why, how and gotchas

#DataBadLuck

86.5%

59.5%

Buying data...why, how and gotchas

#DataBadLuck

Your data vs. your vendor’s

• Your data is less complete

• Your data has a better competitive

advantage

• Use their data to fill in your data

#DataBadLuck

My data is awesome!

MYTH

#DataBadLuck

Points Your scoreFactors 4 3 2 1

Fresh <30 days <60 days <90 days <180 days

Accurate 95.00% 80% + 70% + 60% +

Multi-venueAll

availableBasic + 2 social Basic + 1 social

Basic(email+phon

e)

Built fast <14 days <60 days <90 days <180 days

Normalized Enforced Plan + culture Has plan no

ScoredCustom

rulesAccessible

ruleswhite box

scoringblack box

scoring

Total data quality score:

CRM Data Quality

#DataBadLuck

Points Your scoreFactors 4 3 2 1

Targetedtarget by self description

hand built keywords SIC code

Custom built on-demandmashed from many sources

pulled from larger sample

Complete 95%+ 80.0% 60.0% 40.0%

Exclusive no competitors limited accessanyone can buy

accessfree

TransparentSources

transparentsources known

sourcesavailable

Verified By a personMarketing

automationemail

Total competitive advantage score:

CRM Competitive Advantage

#DataBadLuck

12

24

0 12 24

Data Quality

Co

mp

etit

ive

Ad

van

tage

Where is your CRM data?

#DataBadLuck

12

24

0 12 24

Quadrant Key

Data Quality

Co

mp

etit

ive

Ad

van

tage

Qualitative /Event-Driven

Influence

Quantitative/commodity

Cold Call

newCRMlead

Quantitative /Cyclic

Warm call

QualitativeCyclic

Relationship

CRM+90 days

CRM+180 daysCRM+

360 days

What is your data potential?

marketing

automation

#DataBadLuck

Preventing Duplicate Records Based on Email is Sufficient

MYTH

#DataBadLuck

Preventing Duplicate Records Based on Email is Sufficient

• No. Not even for sending emails.

• Email addresses are not social security

numbers

• True story: I had four email addresses at

one company

#DataBadLuck

My sales team lets me know what they need

MYTH

#DataBadLuck

Data Information Knowledge Process

I want...More data (lists)I want... Better selection (databases)

I want... More contacts per company (zoom)I want... Fresher contacts(Jigsaw)

I want... More information (LinkedIn)I want... More knowledge (many sources)

I want... More process (crm)I want... Sustainable process

The Evolution of Sales Desire

#DataBadLuck

Questions?If you have questions or would like to begin a free trial, please contact us!

Follow us on Twitter

@ringlead

Join the RingLead Usergroup

RingLead.com/LinkedIn

Become a Fan on Facebook

RingLead.com/Facebook

Subscribe to our YouTube Channel

RingLead.com/YouTube

[email protected]

www.RingLead.com@iDonato@michaelforce