Test automation is omnipresent these days. Still, many organisations struggle with implementation. What tools do you select, do you allow teams to pick their own effectieve solution, or do you strive for a more cetralized approach. The answer lies in a carefull balance, where you allow new fit for purpose solutions to emerge, but reduce wildgrowth in the tool landschape.
This presentation explains how we deal with testautomation at DeltaLloyd. In our different departments created working solutions, that are bundled in a Delta Lloyd broad vision on test automation.
In our development we have , ie the test manager of Delta Lloyd with the external consultant , looked at how we could align supply and demand. We used the analogy of the fruit basket. Fruit represents the various types of test automation solutions. With business drivers as a starting point, we did prioritze de development of test automation solutions, use piloting to test the solutions in practize. When a pilot is finished susessfully it was made availble for the the other departments. The fruit is ripe and IT managers can pick it from the fruit basket, knowing that implementation will be easy and swift. The central organization provides solution, knowledge and support.
The presentation will eloborate on the model. How does it help to define the fruit, and the support towards other departments. The presentation covers a wide range from tools, via required skills, resources & processes, upto the aligment with the business. For this we define 4 groups of people, the wholesale, gardener, auctioneer and Consumer, each with its own goals and skill set.
In our presentation we want to share our approach. It benefits Delta Lloyds test automation and surely can help other companies as well.
Survival techniques for testers, beyond the T-shape tester
Experience Story: Implementing Test automation in your organization
1. 9
april
2014
Test
automa3on
in
your
organiza3on,
put
it
in
a
basket
Ard
Kramer
Derk-‐Jan
de
Grood
2. 2
A
shared
experience
Ard
Kramer
• Testmanager
Delta
Lloyd
Leven
• @ard_kramer
Derk-‐Jan
de
Grood
• Principal
consultant
Valori
• @DerkJandeGrood
• hCp://djdegrood.wordpress.com/
3. Mission
Statement
We
want
effec3ve
usage
of
Test
Automa3on
throughout
the
organiza3on
3
Convert
Manual
to
Automated
TesLng
Re-‐usage
of
knowlegde
and
standarized
approach
Tool
selecLon
should
lead
to
leading
tool
8. 8
Forces
and
expecta3ons
TA
Group
Clear
decisions
TransiLon
to
operaLonal
use
Lower
iniLal
threshold
Embedding
in
organizaLon
9. e
Goals
(ask
de
business)
Quality
Efficiency
Time
to
Market
10. 10
Middelen
Tastbare
hulpmiddelen
zoals
tool,
templates,
manuals,
etc
Oplossingen
ImplementaLe
van
middelen
teneinde
een
doel
te
behalen
Dialoog
Zoektocht
naar
de
juiste
oplossing.
Het
bij
elkaar
brengen
van
doelstelling
en
oplossing
Doelstelling
Businessdoel
zoals
gedefinieerd
door
de
business
en
IT
managers
Groene
appel:
Oplossing
die
reeds
ontwikkeld
is
in
andere
divisie
Rode
appel:
Niet
bestaande
oplossing
Time-‐to-‐
Market
Efficiency
Minder
problemen
in
PRD
Businesscase
Businesscase
Businesscase
Consument
Veilingmeester
Tuinder
Groothandel
Implementa3e-‐fasen
I:
Oplossings-‐
gedreven
II:
Vraag
gedreven
III:
Oplossing
en
vraag
gedreven
InteracLe
Aanbod
gedreven
Vraag
gedreven
19. Are
you
ready
for
the
real
challenge…
19
Economische
haalbaarheid
Technische
haalbaarheid
Organisatorisc
he
haalbaarheid
20. 1. Welke applicaties komen in aanmerking
2. Welke Divisies willen graag aan de slag
met test automatisering
3. Opstellen prioritering (roadmap met daarin
TI, PoC en Operationele fase ingepland
per divisie
4. MT besluit ter consolidatie
5. Inplannen TI sessies
Economische
haalbaarheid
Technische
haalbaarheid
Test
Analyst
Test
Navigator
ART
team
Test
Navigator
ART
team
Support
Operationele fase
Test
Analyst
Businesscase
MT sturing
Eerste inschatting
door IT/Test
manager
Commitment op
prioritering en roadmap
Organisatorisc
he
haalbaarheid
Fine tunen BC op
basis van PoC
bevindingen
Hamerslag
operationele
launch
Support
door
ART
team
PoC (proef) faseVerkenningsfase
PoC
rapport
Test
Navigator
In
divsie
Test
Navigator
ART
team
Super impose does not work…
Still management wants to superimpose 1 tool.
There is not one solution that fits all problems
The ideal tools may vary over time and in each project
Introduction:
Tooling starts wiht big story, perfect tool, solution for everything, but in real it takes a year to get a little someting into operation.
Deltalloyd:
TA started on small scale, decentralized. Managing to large skill.
This slide:
Centralized: Big Entetprise tool for all. KPN: QC for all testers. Is this a succes. Yes, tooling availble for every project, No: no control over what is done with it, how it is used, no commitment. (it as option)
Question: does the advantage outweight the investment ?
Decentralized: each project its of solution.
Two extremes:
Many tools, free tools,
e.g. Sticky keys example
Disadvantage: No continuity, maybee more licentce cost, no sharing, not scalable
Other side of the spectrum:
All is regulated.
Ourt expereince is that you need to listen to the teams, they rule the need.
If you it right:
People come look to your garden…how does he do it
People want to learn and benefit from your knowledge
(link with
Succes is determined by people
Oil stain model : shange small, quick and visible)
The forces in the orgainsation
Management likes firm decisions and aims for one tool, one approach
1b) Management has a high over set of goals, that are hard to measure and do not releate to TA in a one to one relation
2) Value is found in using TA and learning to use it
so we need to
lower the treshold (get people to want to try TA
Help them to start using it operationaly
Make it sustainable
Ideal SOLL situation
Who is using tools
Transition from tools to
The users
The things they do with it
So driven by need
Learning there is already a lot of knowledge in the organsiation
Rememer we wanted to lowr trashhold….
We believe this can be done by reusing knowledge and processes…therefor there is need voor repeativeness…preferred set of tools.
No need for BC (other drivers…..?,
MT seems to be convinced that we just shoud do it)
First show results, postpone support organisation until there is a need (
(links with small, fast, visible)
Linke with Change is in people, make them aware of the need first)
This is why the fruitbasked did not got accepted. The model is still used by us, but not understood by our enviroment.
True story, but not the right way of communicatie
This contained a message wch we misunderstood: Maybe the organsiation is not a mature as we think it is. This showed later
Insight: people know what they want….
A focus shift from tools to peole and needs….made us not to SELL but aling with the needs…they people found us…
They were enthousastic about the solution, more than about the tool
Main message: Sales was pushing the solution, but we noticed they cam to us with their needs.
Install some soort of Quality gate
Implemting in the organisation was more to do than just the tool. E.g.
Organisatorische haalbaarheid
Business case
Mangement involvement
Include management…who makes the decisions to do TA
Link with earlier remark: Organisation might be less mature than you think…
Implemting in the organisation was more to do than just the tool. E.g.
Organisatorische haalbaarheid
Business case
Mangement involvement
Basket is tool oriented solution….
But succes is deterimined by other factors…
Basket is tool oriented solution….
But succes is deterimined by other factors…