In Scaled Agile Framework

Actual is a relative term when it comes to business value delivered by a SAFe PI Objective. We had a discussion about this a couple of weeks ago in an Implementing SAFe class and I promised a blog post about this. Here it goes.

UPDATE: I’ll be talking about this concept and related Business Agility themes at the SAFe Summit 2019 in San Diego 

Planned Business Value – Making sure Business Owners and the Agile Team are on the same page

Let’s start from the basics though. PI (Program Increment) Objectives are used as a “back briefing” mechanism by Agile Teams on an Agile Release Train to share their plan for the PI and validate that they are indeed focusing on the highest priorities and are planning to deliver objectives that will be valuable for the business.

Business Owners (Business Executive, Product Management, System Architect) circulate between teams during PI Planning and score each PI Objective on a scale of 1..10 based on business value they ASSUME will be delivered in case this PI Objective is accomplished in the PI.

This becomes the “Planned Business Value (BV)” for that objective.

Actual Business Value – Assessing Business Value based on Demo of a real solution

Later on, during PI Inspect & Adapt (or earlier during System Demos) the same Business Owners circulate between the teams and score each of these PI Objectives again, this time on “Actual Business Value (BV)”.

What does Actual mean here? Well, in most cases the evaluation is based on seeing a demo of a working solution and still making ASSUMPTIONS about what the actual value would be when the solution meeting this objective will be released and available to real users/customers.

Sorry, but while being much closer, that’s still not ACTUAL business value.

ACTUAL Actual Business Value – Based on released solutions and the outcomes they deliver in the real world

ACTUAL business value delivered can be evaluated only AFTER the solution is released.

On most Agile Release Trains / SAFe contexts the PI I&A is too early to make this evaluation so you could understand why we’re still making assumptions at that stage.

But if we really care about outcomes and delivering value, we shouldn’t close the books on these PI Objectives and the PI at that point. We should get back to it later on and Inspect and Adapt based on real business value delivered.

Adjusting the SAFe Inspect & Adapt to track ACTUAL actual Business Value

If you’re with me, you’re probably asking what can we do about this? What is the right timing to get back and assess the ACTUAL actual Business Value? From a cadence perspective there are two main ways to do this.

The simplest is to take advantage of the Inspect and Adapt PI System Demo to review the ACTUAL business value delivered by the objectives in the PREVIOUS Program Increment. E.g. if we’re now finishing PI 5, we’re assessing actual business value delivered by the objectives in PI 5 that will be released sometime during PI 6, as well as PI 4 objectives that hopefully got released during PI 5.

For each one of these PI 4 objectives, now should be a reasonable time to talk about things like – Have customers started to use this solution? Are they happy with it? Did it achieve the impact we had in mind for it? Did we stop incurring the cost of delay we had in mind when prioritising this? At this point, the 1..10 score should be data/evidence-driven.

If we AREN’T able to evaluate the actual business value at this point – that means there’s a short-circuit in our empiric feedback loop that we should work on fixing.

If we haven’t released the solution yet, then we should keep the actual score for this objective empty and get back to it in the next PI. This objective should still be “work in progress” from our perspective.

It’s not DONE until we evaluated the ACTUAL Business Value

You might guess what’s the next aspect of this. Mentioning Work in Progress should be an obvious clue. The Program Kanban has a role in helping us out here as well. Features on the Program-level Kanban shouldn’t be considered DONE until we collected this feedback and evaluated the ACTUAL actual business value on them. They should hang out on the board – maybe in an area called “Feedback” or whatever you prefer.

I’ve been recommending this sort of Program-level Kanban Board structure for a long time now. Some of my enterprise-level clients have improved their Product Management practices dramatically through the accountability and follow-through that this practice encourages.

Just think about the impact on Product Management, Business Owners, Sales people asking for features, if they know they are accountable to the outcomes from these features after they’re released.

Who’s accountable for delivering the actual business value?

This brings us to an interesting question. Who’s accountable to delivering actual business value? Who’s accountable to delivering ACTUAL actual business value? Is it the Agile Team? Product Management? Business? Sales?

I’ve seen way too many teams frustrated when they deliver the objectives according to what they presented as the plan, and yet the actual business value score is lower than the plan because the Business Owners don’t think as much value will be actually delivered. When we’re moving from assumed actual to actual actual the gap can be even bigger. On one hand, in the spirit of transparency and being focused on value and outcomes rather than output, this is the right way to score the business value. It’s about value delivery rather than tracking to plans. On the other hand, you can probably understand the frustration here.

The way I see it, the only way out of this is to understand that the PI Objective plan vs actual vs ACTUAL isn’t an indication of the individual performance of either one of these roles. It’s an indication of the performance of the whole development value stream including the upstream activities related to choosing and prioritising features and the downstream activities related to selling the solution, convincing users/customers to use it,  implementing it in the field, operating/supporting it.

That, together with Lean/Agile Leadership that emphasises principles such as Assuming Variability, Objective evaluation of working delivered systems, and relentless improvement of the whole value delivery cycle, is the key to focusing on learning from these surprises whether they are systemic and repeating or rare exceptions.

A relentlessly improving organisation would inspect what’s the trend when it comes to plan vs actual vs actual actual for the whole program and per specific PI Objectives and try to see what it can learn from when the value gap happens and does it happen for a specific type of objectives or in a specific area of the program/business.

It’s all about Value

Value is the goal of Lean and the fast delivery of value is the goal of SAFe. If we’re serious about that, We should raise our game when it comes to managing value as a first class citizen in SAFe. Business value on PI objectives is the perfect place for doing exactly that.

So, Next time you have PI Inspect & Adapt, don’t just look at the PI you’re finishing just now, take a look at the objectives from the previous PI as well. And on your Program-level Kanban only consider features done after evaluating actual business value delivered, ideally based on quantifiable metrics.

I love it when discussions at class drive me to write up some of my experiences, tips and tricks for the blog. Awesome kudos to my students, now SPCs off to implement a healthy and value-oriented SAFe in their organizations!

Recent Posts
Showing 2 comments
  • Atanas Tasev
    Reply

    Interesting article, I like that people are bringing those topics up in blogs like this one as Scaled Agile doesn’t exactly provide the needed insight on how to tackle some real life questions.

    Disclaimer I am a SPC4 with about 9 months and 1 transformation experience only and PO/PM & SM validated.

    However I’m not fully sure about what you are stating above, that the Actual Business Value is a score you are giving to the traction on the market the Feature delivered achieves. In training materials of the PO/PM course and in the LPM course there is a constant nagging about moving away from lagging indicators to hypothesis validation, so simply based on this I would say it is wrong to keep a Feature “incomplete” for one PI more to validate those lagging KPIs and their expectations.
    Another point is what the Quantitative measurement is actually used for in SAFe for the ART and the Agile teams within it, the Predictability Score, or how much value the teams have delivered and can deliver on average. Now I do believe there is room for interpretation in terms of what “Actual Value” means, but taken from the Product Development side it should be simply if the Benefit Hypothesis is validated fully or partially as at the end that metric should be driving the teams to make decisions about scope locally and then decide if de-scoping this or that might be better than not delivering something else at all.
    Where from Business side the Actual Value is something I believe should be a totally separate metric, maybe on Portfolio level but then again we are not talking about Features but rather Epics where once more SAFe is against trailing indicators.

    I would be interested in debating this topic even more in a more direct communication channel as I do believe it is not exact science and such topics evolve only based on meaningful debate and combined experiences of people in the field.

    • Yuval Yeret
      Reply

      Thank you for your comment and thoughts Atanas!

      Yes we want to validate our benefit hypothesis as quickly as possible. I’m not advocating “keeping” the objective open unnecessarily. what I AM advocating is that if we’re not able to validate the hypothesis, let’s not fool ourselves by marking the objective as “done”, “actual value delivered” and forget about it. Let’s get back to scoring the actual realized business value or validated hypothesis as soon as we can. (like we want to validate business value being delivered during the implementation PI rather than wait until the PI System Demo to evaluate it). Should it be a separate value? Maybe. I don’t think so. We teach SPCs/PMs/POs/BOs to score the business value on objectives according to actual value being delivered (outcome), not fit to requirements (output) so it makes sense to follow up on that at this level as well. Doing this follow up at the portfolio level WOULD be too much of a lagging indicator…

      My main point is that you cannot really validate an hypothesis just by doing system demo from staging. Business Agility is about going all the way to your real gemba of users/customers and seeing whether people actually use the new abilities and enjoy/benefit from them.

      It’s definitely not an exact science – I’m mainly trying to get people thinking about this. Regarding direct communication, I’ll be talking about this at the SAFe Summit in San Diego – looking forward to meeting practitioners and having interesting discussions about the topic during and after the session!
      (https://global.safesummit.com/sessions/validating-delivered-business-value-going-beyond-actual-business-value/)

Leave a Comment

Start typing and press Enter to search