Cascade hell is a real phenomenon, and if you've worked at any company larger than about eighty people that takes OKRs seriously, you've lived through it.
It starts cleanly enough. The leadership team sets three company-level objectives. Those get handed to the VPs, who break them into department-level objectives. The department-level objectives get handed to managers, who turn them into team-level key results. The team-level key results get handed to individual contributors, who write personal OKRs that are supposedly derived from the chain above them. Six weeks later, an engineer is staring at four "personal key results" with numerical targets she can technically trace back to a company objective if she squints, but which she experiences as a list of tasks she's been assigned. The strategic intent that started at the top has been translated, paraphrased, and reformatted four times. By the time it reaches her, it's not strategy anymore. It's chores with a numerical wrapper.
That's cascade hell. And it isn't fixed by getting better at cascading.
Why the cascade degrades
Every layer of translation introduces interpretive error. A company-level objective gets read by a VP whose context, incentives and recent conversations colour how they translate it. The VP writes a department objective that reflects their interpretation. A manager reads that and translates again — adding their own context, their own incentives, their own constraints. By the time it reaches the front line, the original intent is unrecognisable, and worse, nobody at the front line has the context to notice that it's been distorted.
There's a second problem on top of the translation problem. Cascades produce volume. If three company objectives become twelve department objectives become forty team objectives become two hundred individual OKRs, you don't have alignment. You have a documentation industry. Every person in the company spends a week each quarter writing OKRs nobody outside their manager will ever read, in service of a strategy whose connection to those OKRs nobody can audit.
If you removed the personal OKR layer entirely, would your company execute worse, or just have less paperwork? Most leaders, asked privately, will admit they're not sure. That uncertainty is the diagnosis.
Contribution mapping: cascade in reverse
The alternative is to invert the direction. Instead of breaking goals down from the top, teams map their work up to the top. The mechanics look superficially similar — there are still teams, still goals, still company objectives — but the underlying logic is different in a way that matters.
Each team answers two questions, in their own words. Which company objective does our work contribute to? What specific outcomes does our work enable that serve that objective? The answers don't come from a template handed down from above. They come from the team's own understanding of what they do and why it matters.
This is harder than the cascade. The cascade is arithmetic — break a number into smaller numbers, assign each smaller number to a team. Contribution mapping is judgement — the team has to understand the strategy well enough to articulate, in their own language, how their work serves it. That's a higher bar, and a much more useful one, because a team that can articulate the connection has actually internalised the strategy. A team handed a cascaded KR has only been handed a target.
The other thing contribution mapping does is surface the gaps. When five teams claim contribution to objective A and zero teams claim contribution to objective B, you've discovered something. You've discovered that objective B is either implicitly nobody's job, or so obviously someone's job that nobody felt they needed to claim it — both of which deserve a conversation. The cascade would have hidden this by mechanically assigning fractional ownership to everyone.
Four steps to do it
Step one is to share the company objectives with full context. Not just the bullets on the slide. The trade-offs that produced them, the logic that connects them, the bets the leadership team is making. Teams can only map to a strategy they understand. If the strategy has been communicated as a list of targets stripped of reasoning, the mapping exercise will produce a list of targets stripped of reasoning. Garbage in, garbage out.
Step two is to ask each team to name their contribution explicitly. For each company objective: does your team contribute to this? If yes, how, specifically? What outcomes does your work enable? Write it down. Be specific. Vague answers ("we support the company's growth ambition") should be pushed back on until they become concrete.
Step three is to review the gaps and overlaps as a leadership team. Objectives with many claimed contributors often have a diffusion-of-ownership problem — when everyone is contributing, nobody truly owns. Objectives with no claimed contributors are underfunded, full stop. Both findings deserve a redesign, not a footnote.
Step four — and this is the one most companies skip — is to make the contribution visible in the place the work actually happens. A contribution map that lives in a slide deck is a contribution map that will be forgotten by week three. The connection needs to be embedded in the system where work items get created, so that when a team member opens a new task, they can see which company objective it serves without having to remember a separate document.
The Vindaris view
The cascade is an artefact of an era when goals and work lived in different tools and could only be connected by translation. When goals and work live in the same system, you don't need to cascade. The contribution map becomes a property of the data: every work item knows which objective it serves, every objective can show every work item beneath it, and the connection between team execution and company strategy stops being a slide and starts being a live structure.
That shift — from translated targets to a visible contribution graph — is what makes "team goals tied to strategy" actually true, rather than ceremonially true. And it's the only way a company of more than a hundred people can keep that connection honest without burying everyone in OKR paperwork.