Nowadays trade within the rising digital and e-commerce driven world, added value through digitization, increasing requirements
for time to market, all of these are movements that put companies under even more pressure to succeed.
Moreover, a customer can often easily and quickly replace his seller. Sometimes the competitor is just a click away.
It is therefore even more important to dedicate to the customer's wishes, to accompany them with high-quality information
in their decision-making process (e.g. user journey, user experience, UX), to guide them with meaningful product information
(i.e. just right amount of information, precise, reliable), as well as to match the required circumstances of different
channels (e.g. online, on-site, print)
This is where
The power of PIM, digital assets, ...
PIM excellence
PIM best practices
Excursus: symptoms if having no PIM or an inferior implemented PIM
The power of PIM,
digital assets, ...
Because of this 'acting behind the scenes', the extremely huge potential of power of data, that can be obtained through
master data environments such as PIM, often only becomes visible when you take a very close look at it.
To start with, PIM is not a self-propelling success (see below for further details, pls). Poorly done, the
risk of failure and 'burning money' is extremely high.
But first of all, find following essential of PIM's
- Central repository with the central
truth of data (i.e. single source of truth) Trusted , reliable, credible, true, consistent, robust, structured, semantically correct, relevant to customers, accurate, uniformed / harmonised, quality-assured and up-to-date data.- Product data aligned with the business purpose for a
holistic customer experience and to support product-related decision-making processes - across the entire range of products offered. - International markets,
multilingualism , localisation. - As a trade company dealing with the
data of many suppliers , frequently in very different formats, structures and quality. - Well shepherded product data, stewardship and onboard new products, including
governance guidelines ,style guides ,compliance with regulatory demands and responsibilities (RACI ) - Well-organised and
strong product taxonomy and hierarchies . Typically multiple hierarchies (e.g. for customer journey, reporting), and separated from PIM internal hierarchy (unambiguous, attribution). - Strive for excellence in case of enormous product offerings with
million of products to offer, too. - Appropriate treatment of the fact that product information are
subject to dynamic change over time. - Move data across different
IT systems and marketing channels (downstream exports, multi-channel, omni-channel, web, print, e-commerce, e-procurement, syndication, variations depending on channels, etc). W/o PIM it is difficult to impossible to share master data assets between the channels, potentially adapt them to varying needs, as well as to orchestrate all this reliably. - Operational excellence through links to many digital assets (like
high-definition images , product image zooms, short product videos, tutorials, all of this potentially in different languages, and the like) - Facing a world of
dynamically changing product information.
Finally, let me briefly explain what is meant within the headline 'PIM, digital assets, ...' by the enhancement and the dots. In case it is treated properly, PIM Product Information Management is undoubtedly at the heart of achieving the above listed features and ultimately an important enabler for digital growth. Nevertheless, depending on the industry, competitive situation and existing / legacy IT infrastructure there might be a fine-tuning with related systems. These can typically be: PLM (Product Lifecycle Management), MDM (Master Data Management) and/or DAM/MAM (Digital/Media Asset Management).
PIM excellence
The introduction of PIM within a company should not be confused or equated with an introduction of a tool only.
Such a PIM undertaking is almost always
.jpg "Knut Höngesberg")
This results in quite extensive requirements for the expertise and competence of a
- In-depth knowledge of
information modeling at the business level, levels of database normalization, standardisation, logical data modeling at the entity-relationship level (ERM / ERD), etc. Best case, also understanding and experience with data modeling at the level of the technical data model, which additionally takes e.g. performance aspects, treatment of redundancies on purpose, etc into account. - Know-how in
business process modeling (BPM) and business process design, using e.g. BPMN or ARIS event process chains (EPC) Analytical skills and associated methodological competences (e.g. SWOT, C&E cause & effect matrix, RACI, ITSM / ITIL)- Deep insights and practical experience in
operative master data management (i.e. have gained a solid gut feeling for what it means to domesticate something that appears like a 'flea circus'. Btw, a 'flea circus' that initially and typically seems to materialise from nowhere, and not from an event where someone did something wrong) Authenticity , like being able to carry out every step of a final implemented PIM / MAM / DAM (digital assets) environment yourself that you would expect from others (e.g. from a data steward, PIM attribute owner, or downstream from marketing people)- Knowledge of common
standardisations for master data exchange / syndication (e.g. GS1, 1WorldSync, BMEcat XML, eCl@ss, ETIM, UNECE, GDSN, EDI EANCOM PRICAT, etc) - Knowledge of common
PIM, MAM/DAM, MDM tools (e.g. SAP Hybris, Informatica 360 / Heiler, Novomind iPIM, JCat, Contentserv, Riversand, Akeneo, Censhare, Celum, SAP MM, etc) - Comprehensive
project management and project lead experience . Due to the characteristics of PIM ideally both agile (e.g. Scrum) and conventional (waterfall) - And last but not least:
very strong interpersonal skills .
.jpg "Knut Höngesberg")
- More then 15 years of PIM experience in leadership roles.
- B2C and B2B. Very successful early adopter of B2C market implementation powered by PIM.
- Accomplishment of several very comprehensive, huge and challenging PIM environments. Furthermore, a couple of times deep insights into poor examples (i.e. how not to do it) (P.S.: the poor ones were not led by Knut into their stage)
- Project management covering time, budget and quality / scope. Thereby using agile as well as conventional methods.
- Internationality, especially Europe / EMEA and USA.
- Well connected within PIM industry / service providers in Germany, Europe, US, India, Latvia, near-shore and off-shore.
- Cooperative management style as a guiding principle. Appreciative collaboration including team spirit, team motivation and team coaching are seen as key success ingredients of a well-functioning and innovative digital team.
- PIM rainmaker: loves to build
hyper-competitive and customer-obsessed digital business successes, with the ability to execute, and out of a 'play to win' teamwork.
PIM best practices
In case you might want to try it yourself (potentially as a start) and without the help of an expert like me, find following some tips and hints for you on what to take care of:
- Extensive involvement of all related people (technical and non-technical users), decision-makers and stakeholders at all
relevant organisational levels. Of course, it sounds naturally, but make sure you really live it. PIM & Co is
hard work of people and processes , and extraordinary much driven by a high team spirit, combined with strong and reliable connections to organisational stakeholders. - Coordination and planning around
data governance , governance policies, style guides and the like are crucial. - PIM appears like millions of 'little issues', each of it frequently not even worth enough to talk about.
Therefore,
abstract complexity , work (also) consistently on the 'big picture' as well as on the completeness of vision. Think not only in projects, but in programs, and of course, obviously with the customer in focus. I recommend that you tackle the topic based on a program bysliding the elephant , agile plus sustainable, and with many milestones of business successes on the roadmap. - Build a
logical (i.e. business) data model (e.g. with the representation of product-variant relationships, elements of configurable products, etc), normalise it, and ideally also draw it. Considering that, use e.g. methods of information modeling or entity-relationship-diagram (ERD), including one-to-many expressions. At first impression, such may look a bit technical for business people, but from my point of view it is extremely helpful if non-technicians learn to recognise their product data structures. It may only need a bit of guidance to introduce it. Resist pure belief in tools . Technically the PIM Tools can do many things in many ways, but if you leave everything open or use everything that is technically possible, uniformed structures might not arise, and you will not get your product range comparable or filterable to serve your customers and prospective customers needs.- Robust
attribute modeling , consistent attribute readiness, andinheritance mechanisms (media neutral). - Pay attention to well-designed
units of measurement (UOM) as well as the concept for the units. - Setting up
metrics andkey figures to measure success, in particular for measuring business successes. I don't believe it will come to the point where you think you have 'enough' data quality, or the topic is 'done'. Master data like PIM are always a rolling stone and a continuous topic for further improvement. It needs key performance indicators (KPI), triggers for right prioritisation and measurable goal setting. At the same time, you gain a profound basis to fulfillservice level agreements (SLA ). - Strong focus on business requirements and use cases, business processes and data that drive sales and business success (ROI).
Keep an eye on the entire business process chain, full
end-to-end productiondownstream to the customer. Furthermore, automate processes where feasible. Typically, these are always monotonous tasks that nobody really enjoys doing manually. Take the 'low hanging fruit ', it's worth it. - Develop, establish and train responsible
organisational structures , e.g. theownership of PIM processes and data model (information model), the PIM attribution, the operational stewardship, etc - Build strong and
well-organised product taxonomy and hierarchies . Define and recognize the application purposes (example: for the PIM-internal hierarchy including the attribution, uniqueness is extremely important. Instead, in the hierarchy for the customer journey in a shop the opposite is typically desirable). In case you support several product hierarchies in parallel, make sure that the effort remains manageable, especially for the product assignment. In such context, to boost efficiency I like very much (partial) mappings and inheritance between different product hierarchies. - Proper analysis and concept of
data migrations from legacy databases , files, spreadsheets, etc into the PIM landscape. Think about the different possible data migration scenarios, e.g. as a 'big bang' (which is of course convenient, but often unrealistic in operational implementation), or 'iteratively with data quality gates'. I recommend developing data migration concepts already in parallel at early project phases. It is incredible positive how much you can learn iteratively from your existing data towards your new information model. - Think about the
future , future challenges and demands, it is sure to come. Avoid dead-end approaches. Flexibility within PIM (and even master data management in general) can frequently be gained with a snip of the finger. It's more about stayingmethodocially clean and doing it right the first time. - Wherever possible,
industry standards in data exchange should be used (e.g. GS1, UN, BMEcat XML, eCl@ss, ETIM). Such helps every partner, including you. Both are of interest, i.e. - if you are a dealer - on the data input side (e.g. content transfer from your suppliers or service providers into you databases), as well as on the data output side in the direction of your customers, marketplaces, etc - In case you are a trader, consider whether you should drive a
bonus / malus system based on meaningful KPI's for the data input side. You could track supplier performance as well as quality of data delivery, and reward your supplier to gain market power, live the win-win idea with your supplier, and the like. - If you do business internationally and have to use several languages to reach out to customers, it is typically useful
to embed a
translation memory (TMS) and/or a translation service provider. - Frequently, business processes and tool functions can be used '
out-of-the-box '. The tools for PIM, MDM, MAM / DAM, etc often provide a lot therelike. Obviously, the closer you stay with the standard the easier and cheaper it is to migrate to new versions of the standard tool. - Finally: consider the importance of such best practices, they are often the game-changer.
Excursus: symptoms if having no PIM
or an inferior implemented PIM
The following is a list of frequently seen symptoms (often even in combination), that can occur if your PIM environment is not set up professionally enough or even missing at all. This list may not be exhaustive, but the elements can be taken as typical indicators:
- Customers in your online shop or product catalog cannot find your products easy and reliable. The attribute quality does not support a proper selection or filtering (drop-down lists). Attribute values look partially trashy or useless.
- Awkward behaviour of data that seem to ominously destroy themselves as if guided by a third hand, and/or the source for attributes and attribute values is inconsistent and competing. Potentially, the methods of inheritance were not understood or used incorrectly.
- Product hierarchies (taxonomies) do not work properly and there are contradicting uses of it from different departments. For the PIM-internal product hierarchy (i.e. the hierarchy to drive attribution) sometimes uniqueness is missing (i.e. the assignment of an article X would be meaningfully in both correct, in a hierarchy tree A as well in a hierarchy tree B)
- Data maintenance efforts for duplicated information increase (potentially even exponential), because redundancies within the data structures were created by mistake or for workarounds.
- Activities for data quality improvements do not work or do not work properly. The current data status (KPI) can not reasonably be quantified and doesn't appear to be tangible. You feel no sustainability, maybe even experience a feeling of powerlessness.
- Similar types of data that have only arisen at different times (legacy data, data migration, data from cleaning phases) do not match the data of the same type that are currently being created. Even if there should be same values, they appear sometimes completely different.
- Data movements between systems are unreliable. In various systems, outdated data sometimes reappear unexpected, are partially missing, are inconsistent or contradictory. Resultant even legal issues may appear. All of this leads to loss of credibility in data, and sometimes to the situation that departments feel the need to create local workarounds. Such affects negatively the technical vulnerability, complexity in treatment of content management and unreliability of data movement. So from an overall perspective, everything gets worse (possibly even exponentially).
- You cannot easily reuse your data in different channels (simplified example: within eCommerce channels, you might want the customer to be able to filter your product assortment with attributes such as width and length. On the other hand, in the print catalog you need to save printing space and only want to output width and length as a combination 'measurements' (e.g. '2x5inch') Worst case such is maintained twice)
- Your dimensions and units of measurement do not match, cannot be filtered properly or cannot be converted (e.g. from millimeters to centimeters). Potentially, you have inconsistent data between imperial and metric systems of measurement (e.g. inches and cm), or from the customer's perspective awkward data such as a diameter of '50.0003 inches'.
- You suffer from media breaks.
- In the event that your company could operate in international markets and serves translations of texts into other languages, you might see impacts (especially in the event of subsequent information changes) of inconsistencies across languages, not followed up translations, and/or might face constantly high translation costs. Furthermore, probably a translation management or translation memory (TMS) is missing.
- The list of your business requirements related to product information from e.g. the marketing side is getting longer and longer, w/o having the possibility to implement important requirements quickly enough. Additionally, your business users are facing a lack of flexibility.
- Costs get out of hand / huge budget overruns.
If some of these aspects might frighten you and/or you partially might recognize your company like in a mirror,
please do not be discouraged. The good part of the news is that all of this can definitely be solved with logic,
dedication to the topic, and strong teamwork. In case you follow the recommendations from above, plus - potentially -
bring in the right coordinating person into your company, a lot has already been won.
I wish you every success as well as fortunate decisions at all times, to build within your company a customer-obsessed
environment powered by PIM, to be able to implement your digital strategy successfully, and to shape your digital future
to win business.
If you have any questions, please do not hesitate to contact me. If feasible under potential competition constraints,
I will be happy to help you.
The electric light did not come from the continuous improvement of candles.
(Oren Harari)
