Las operaciones de marketing son la parte del departamento que nadie fotografía para la web corporativa. Son la fontanería (los flujos de trabajo, las tuberías de datos, los modelos de atribución), y cuando funcionan, el aplauso se lo lleva la campaña que casualmente estaba corriendo por encima. Sentamos a Mira Wise para que nos describiera cómo es de verdad la sala de máquinas, y llegó con la calma pausada de quien ha visto fallar un píxel de seguimiento en tiempo real y ha sobrevivido para reconstruir los datos después.
Entrevista Expert Insight


Wise pasa aproximadamente el ochenta por ciento de su vida laboral dentro de HubSpot, una herramienta de la que depende y a la que detesta, con frecuencia en la misma frase. Lo que sigue es un recorrido guiado por los sistemas, las averías y la política de oficina que deciden si un equipo de marketing puede creerse el número de su propio dashboard, o si en cambio pasará una quincena discutiéndolo.
Conclusiones clave
Conclusiones clave
- El techo de HubSpot: HubSpot mueve todo el stack de go-to-market justo hasta el momento en que ya no puede, y Wise choca de forma rutinaria con el punto en el que la herramienta simplemente no hace lo que necesita, enviando el trabajo a Tableau o a una hoja de cálculo.
- La alineación gana a la ingeniería: Pasar a una atribución avanzada nunca fue de verdad un problema técnico. Lo difícil fue conseguir que ventas, marketing y partnerships se pusieran de acuerdo sobre qué significa la palabra “crédito”.
- El tráfico directo mentía: Una porción sospechosamente grande de “tráfico directo” resultó ser un seguimiento entre dominios roto. Dividirlo en directo real, bug y consentimiento rechazado recuperó la confianza de la dirección más rápido que cualquier dashboard.
- La atribución perfecta se acabó: Entre las normas de privacidad y el descubrimiento impulsado por IA, Wise espera que el marketing mida tendencias y uplift en lugar de touchpoints precisos, y cree que algunos equipos siguen en la fase de negación.
El ochenta por ciento de la jornada transcurre dentro de HubSpot
Existe una fantasía, popular en las conferencias, en la que el profesional moderno de operaciones de marketing preside una resplandeciente constelación de herramientas best-in-class, cada una zumbando en perfecta integración. El martes real de Wise es bastante más monógamo que eso. Su día empieza y termina en HubSpot, que ejerce a la vez de plataforma de marketing automation y de CRM en todos los equipos de go-to-market. A su alrededor orbita un reparto secundario: Customer.io para la comunicación con clientes, Zapier y n8n sosteniendo el middleware, Mixpanel y Tableau para el análisis que HubSpot no alcanza, y Slack, que ella describe, con la franqueza de la trabajadora permanentemente remota, como su cordón umbilical.
La relación con HubSpot no es tanto un romance como un matrimonio largo y con opiniones. Valora la interfaz, la velocidad a la que la empresa lanza funcionalidades y el alivio de no tener que cuidar de una sincronización entre Salesforce y Marketo. Lo que no finge es que la herramienta todo-en-uno se mantenga robusta en los bordes.
“Es curioso, es a la vez la herramienta de la que más dependo y en la que más tiempo paso, pero también, por eso mismo, a veces es la herramienta que más me frustra.”
El patrón se repite las veces suficientes como para ser una ley de la naturaleza: empujas a HubSpot hacia una tarea de nivel enterprise, chocas con un techo y el trabajo se traslada en silencio a una hoja de cálculo o se construye desde cero en otra parte. Lee nuestro análisis completo sobre el stack de herramientas diario de operaciones de marketing.
La pregunta sobre el dashboard que nadie hace primero
La mayoría de la gente cree que un proyecto de dashboard empieza construyendo el dashboard. Wise empieza dos pasos antes, con una pregunta que suena casi grosera de tan simple: quién va a mirar esto, y qué quiere medir en realidad. Un CMO y un campaign manager asentirán ambos con entusiasmo ante la frase “dashboard de marketing de funnel completo” y luego querrán cosas completamente distintas de él, a granularidades completamente distintas. Sáltate esa conversación y habrás construido un único artefacto que no satisface a nadie, lo cual es una forma singularmente desmoralizadora de gastar un sprint.
Solo una vez fijados la audiencia y la definición de éxito pasa a la parte mecánica: confirmar que el dato existe en el sistema, traer lo que falte, hacer QA de los valores y elegir visualizaciones que no engañen activamente. El orden importa. El dashboard es lo último que se construye, no lo primero.
Y la métrica que usa para juzgar si todo el ejercicio funcionó no es una métrica en absoluto. Es la confianza. Cuando los stakeholders dejan de preguntar si los números están bien, dejan de preguntar cómo funciona la cosa y empiezan a usar el dashboard sin hacer ruido para tomar decisiones, el proyecto ha triunfado. Un dashboard precioso que nadie abre es, en términos de operaciones, un fracaso con buena producción.
Ingresos, o no contaba de verdad
Las métricas de funnel son seductoras porque llegan pronto y parecen progreso. Una campaña generó tantos leads, tantos SQLs, y la tentación es dejar de leer ahí. Wise lo trata como el primer capítulo de una historia, no como el final. Los equipos a los que da soporte intentan, allí donde el seguimiento lo permite, rastrear cada campaña hasta los ingresos que acabó produciendo.
La razón es un escenario que ha visto repetirse. Dos campañas corren en paralelo. Una convierte a SQL al veinticinco por ciento, la otra al diez, y la conclusión obvia es que la primera campaña es la ganadora.
“Pero si esperas a ver cuál es el resultado de ingresos de esos clientes y te das cuenta de que ese 10% nos está trayendo de media mucho más ingreso por cliente, eso te cambia un poco la mentalidad sobre dónde vas a apostar de verdad tu dinero.”
Ese es el argumento entero a favor del reporte anclado en ingresos en dos frases. Las tasas de conversión halagan a la campaña equivocada. El ingreso medio por cliente te dice adónde debería ir el dinero. El seguimiento nunca es lo bastante perfecto como para hacer esto fácil, pero cuando el dato lo permite, es lo primero que mira la dirección.
El modelo de atribución que HubSpot se negó a construir
La mayor queja de Wise con HubSpot es concreta y, para cualquiera que haya usado Salesforce, inmediatamente reconocible. HubSpot no tiene un objeto campaign member. No existe un registro individual que vaya anotando en silencio cada touchpoint del recorrido de un contacto: el registro al webinar, la solicitud de demo tres meses después, la fuente de cada uno. HubSpot ofrece first-touch y last-touch, lo cual es mejor que el único campo de atribución con el que su empresa empezó, pero deja a un equipo de marketing ciego ante casi todo lo que ocurre en el medio de su propio funnel.
Así que el equipo construyó alrededor del agujero. Una analista de datos levantó un modelo de atribución a medida que vive fuera de HubSpot, en Tableau, donde el dato digital de backend existe aunque la interfaz de HubSpot se niegue a mostrarlo. Eso resolvió los touchpoints online. No resolvió los offline (listas de asistentes a eventos, leads referidos a través de PartnerStack) porque esos nunca ocurrieron en un navegador que HubSpot estuviera vigilando.
La solución fue reconstruir a mano el concepto ausente de Salesforce: un campaign member casero dentro de HubSpot, exportable al modelo de Tableau. La recompensa no es cosmética. Alimenta con cinco interacciones antes invisibles un modelo de atribución ponderado y todos los pesos existentes se desplazan. El modelo refleja por fin a ventas y a partnerships, no solo a marketing. Lee nuestro análisis completo sobre cómo crear un modelo de atribución a medida sobre HubSpot.
El día en que el seguimiento se rompió en silencio
Algunos fallos se anuncian solos. Este no. Tras el lanzamiento de HubSpot, el equipo notó una porción incómodamente grande de tráfico aterrizando en el cubo de “directo”. Que el tráfico directo suba es normal en una era de bloqueadores de cookies, así que la jugada fácil habría sido encogerse de hombros y culpar a la privacidad. Wise no se encogió de hombros.
La investigación descubrió que el seguimiento entre dominios entre los muchos dominios y subdominios de la empresa estaba roto. Un visitante que pasaba del .com al .nl se contaba como una sesión completamente nueva y una conversión directa, cuando no era nada de eso. La visibilidad de campaña se escapaba por la rendija. Peor aún, el código de la web pertenecía a producto e ingeniería (un equipo que trabaja en sprints ajustados con su propia hoja de ruta), así que el equipo de marketing no podía parchearlo sin más.
Lo que convirtió un sprint de pánico en un proceso estable fue la honestidad sobre el tamaño del problema. En lugar de reportar un único número aterrador, el equipo dividió el tráfico directo en tercios: realmente directo, el bug entre dominios en reparación activa, y tráfico que las normas de consentimiento hacían genuinamente imposible de rastrear.
“Es un 25%, no un 100% completo de lo que pensábamos que era, y eso también ayuda a recuperar la confianza, que creo que se puede perder en situaciones como esta.”
Lee nuestro análisis completo sobre cuándo el seguimiento entre dominios se rompe en silencio.
Decirle a un stakeholder que su idea es cara
Las operaciones se sientan en el asiento incómodo entre ingresos, marketing y ventas, lo que significa que absorben un tipo concreto de fricción. La clásica disputa entre ventas y marketing por el crédito es una fuente. La más silenciosa es la petición que llega ya resuelta: un compañero aparece queriendo una automatización concreta o un campo concreto rastreado, habiendo decidido ya cómo debería funcionar, y operaciones tiene que ser quien da la mala noticia sobre lo complicado que es eso en realidad.
Wise no gestiona esto diciendo que no, ni tampoco diciendo que sí. Lo gestiona interrogando la petición hasta que aflora el verdadero problema que hay debajo, y entonces despliega las opciones con sus costes pegados. El factor decisivo son los ingresos, más la alineación con lo que la empresa ya haya declarado prioritario. Una campaña que alimenta un objetivo actual de toda la compañía (su ejemplo es convertir clientes de planes mensuales a anuales) se gana sus cuatro semanas de construcción. Un experimento especulativo no, al menos no a escala completa.
“No vamos a pasar cuatro semanas construyendo algo para una campaña experimental de la que no sabemos cuánto ingreso va a traernos de verdad.”
La elegante válvula de escape es la versión reducida. Construye el experimento pequeño, demuestra el impacto y la conversación con la dirección sobre la inversión completa se vuelve fácil. Lee nuestro análisis completo sobre cómo las operaciones de marketing priorizan las peticiones de los stakeholders.
El final silencioso de la atribución perfecta
Preguntada por hacia dónde se dirige el marketing B2B, Wise no recurre a la euforia habitual con la IA. Su respuesta es más serena y bastante menos cómoda: el seguimiento perfecto se acaba, y un buen número de equipos aún no lo ha procesado emocionalmente. Entre las normas de privacidad y consentimiento y el auge del descubrimiento impulsado por IA, el sueño de capturar cada touchpoint digital con precisión está, en su valoración, expirando en silencio.
Lo que lo sustituye es una filosofía de medición tomada prestada de un lugar decididamente analógico: la campaña de exterior. Nadie le pone un camino de atribución limpio a una valla publicitaria. Miras si el tráfico y los ingresos subieron en las regiones donde corrieron los anuncios, y aceptas la inferencia. Wise espera que la medición digital derive en la misma dirección: rastrea la porción que puedas, reconoce la porción que no, y lee uplift y tendencias a lo largo del tiempo en lugar de exigir una línea perfecta del clic al trato.
“Ahora tenemos que aceptar un pequeño giro hacia más tendencias y uplift que precisión perfecta.”
No está de luto. Sospecha que nuevas herramientas, posiblemente un agente de IA que detecte cambios de tendencia que un humano pasaría por alto en un dashboard, acabarán llenando el hueco. La industria, cree, simplemente sigue en la fase de aceptación. Lee nuestro análisis completo sobre por qué está terminando la atribución perfecta.
Lo que una carrera en operaciones enseña de verdad
Preguntada por la lección que hizo a sus equipos mediblemente mejores, Wise ofreció dos, y ninguna implica software. La primera es entender el ciclo de vida completo del cliente: cómo capta marketing la atención, cómo la convierte ventas, cómo customer success hace crecer y retiene la cuenta. Esa visión de extremo a extremo es lo que asciende a una persona de operaciones de tramitadora de pedidos a socia de negocio, porque te deja ver dónde se solapan los equipos, dónde se aíslan en silos y cómo encaja de verdad la máquina.
La segunda es la disciplina de hacer mejores preguntas. La gente de operaciones, por temperamento, quiere ayudar y quiere agradar, lo que la hace propensa a convertirse en una máquina de decir que sí que construye exactamente lo que le pidieron en lugar de lo que se necesitaba. La corrección de Wise es escarbar hasta que el objetivo genuino quede claro, incluso cuando la petición original era vaga, porque solo entonces el dashboard, el sistema o el proceso puede resolver de verdad el problema. La respuesta del manual y la respuesta que necesita tu negocio concreto no siempre son el mismo documento, y saber cuál seguir es, al final, la mayor parte del trabajo.
Transcripción completa de la entrevista
Read the full interview transcript
Sophie Steffen: Hey, everyone. Welcome to Let’s Talk Marketing. Today we’ve got Mira Wise joining us. She’s marketing operations manager at SendCloud. Marketing ops is basically the engine room that keeps campaigns running, but it’s almost never as clean as the dashboards make it look. Today we’re totally skipping the high-level strategy part. Instead, we’re going to focus on the reality of revenue and marketing operations, meaning how data, systems, and processes actually come together behind the scenes to drive pipeline, and bridging that awkward gap between stakeholder requests and technical reality. Welcome, Mira. Great to have you on the show.
Sophie Steffen: Do you want to give us a quick intro about yourself and what you do?
Mira Wise: Yeah, of course. Thank you so much for having me. My name is Mira. I’m originally from the US but based in Barcelona. At the moment I’m working for a Dutch company, but Barcelona is my home base. Early on in my career I did a little bit of everything in marketing for startups and small businesses. So I was doing paid, social, events, lifecycle. And then eventually that evolved more into marketing automation and operations, so more systems and processes. And then that evolved into more of a revenue operations focused role. So now what I really do is I work with the full go-to-market team, marketing, sales, customer success, partnerships, for B2B SaaS companies, helping them build the systems and the processes that keep their day-to-day running. And also making sure that the leaders of those teams have the data they need to know whether or not their teams are being successful.
Sophie Steffen: Perfect. I love the summary, short and sweet. So now you’re into ops and revenue. Let’s look at a normal Tuesday for you. What exact software did you have open on your screen when you were actually looking at how to keep things running across the go-to-market teams?
Mira Wise: Yeah. So I would say I spend probably 80% of my time in HubSpot. My day normally starts and ends there. That’s the backbone of our tooling for all of our go-to-market teams. We have a couple of different instances, but every one of those teams is essentially using HubSpot. We have the HubSpot marketing automation platform, and we also use HubSpot CRM. So I spend a lot of time there managing workflows and dashboards and reporting. And then I’m working in a lot of other systems that the marketing team is using. So like Customer.io, which we use to communicate with customers via email or app. And all of these middleware tools or AI tools like Zapier or n8n. I spend a lot of time, as I’m sure most of us do now, in Claude and ChatGPT and using AI. And Slack is always open on my screen, especially being remote. Slack is like my lifeline. And then we have reporting systems like Mixpanel and Tableau for deeper analysis.
Sophie Steffen: Do you have any favorite tool? I feel that Slack is the one, but HubSpot is the tool you probably use the most. Is there any tool you can rely on totally, and any other tools where you think, okay, maybe this one will let me down at some point?
Mira Wise: Yeah, it’s a good question. HubSpot is, it’s funny, both the tool that I rely on the most and that I spend the most of my time in, but also because of that, it’s sometimes the tool that frustrates me the most, because we really use HubSpot to its max capabilities, especially as a CRM. HubSpot is really good. It has an incredible UI. It’s very easy to use, very easy for a marketer or a salesperson to use. And it’s also all-in-one, so that’s really nice. If you have HubSpot marketing and CRM, you’re not managing something like a Salesforce and Marketo sync, which can be a big headache. So that’s all lovely about HubSpot, but then there are trade-offs, which means the tool is not always as robust as some of those enterprise systems. So eventually you hit a ceiling of, okay, I’m trying to accomplish this thing and I realize I’ve maxed out what HubSpot can do. I do really like the tool. They’re also very agile and they release a lot of things much faster than an enterprise company typically would. But at the same time, sometimes you realize, oh, I have to build this outside of HubSpot, or I have to build this from scratch because the tool just can’t support it.
Sophie Steffen: Well, not every tool is perfect, but if it gets most of the job done, that’s already a start. Moving into workflows, when you hand over a very pretty dashboard to the marketing team, what does the actual workflow look like to get the data clean?
Mira Wise: Yeah. So when I’m handing over a dashboard, there are a few things I keep in mind. One of those is obviously QA-ing the data, making sure it makes sense, double checking that it’s accurate, and double checking that the visualizations you’re using are the best way to look at that data. But I normally take a step back before we even get to the point of building the dashboard. That workflow starts with who the audience of this dashboard is and what they actually want to measure. You might have something like, we need a full-funnel marketing dashboard, and that’s great, but is a CMO going to be looking at this, or is this also campaign managers? Because those two people care about different metrics and at a different granularity. And then it’s important that either it suits both, or you’re building two dashboards, so that the person you’re building this for gets the high-level insights they need or the detailed insights they need. So I’m typically starting with what does success look like and who’s going to use this. And then you get into the part of, okay, do I have all of the data I need in that system? If I don’t, how do I pull that data in, and how do I visualize it in a way that makes the most sense for the stakeholders?
Sophie Steffen: Yeah, that totally makes sense, to first start with the stakeholder, and then which data points you need and how you get them in there. Is there any step within this workflow that is smoother than others, or any part where you feel, wow, this is always the same part where I have to hustle through things and it’s not as smooth as you’d like?
Mira Wise: Yeah, I think I’m pretty lucky now. We’ve done a lot of work to build up our data infrastructure, especially in tools like HubSpot, and we have a really big data operations team that will help you get everything from our data warehouse that we’d need on a customer into our systems. So we’re pretty good at, if something is missing, I know exactly who to go to and they can help me push that data in so we can build those reports. But there are also always things where maybe you can’t really report on that in HubSpot as reliably, so you do end up going out and looking in a spreadsheet to do a deeper analysis, or going to the data team and saying, hey, I think this is better for Tableau. Something like a multi-touch attribution model, for example, is something I would always choose to report on in Tableau rather than within HubSpot directly. But for the majority of the things we build, we’re pretty lucky that we have a really solid base for all of the data that comes in.
Sophie Steffen: Cool. Going a little bit deeper into workflows and also pipeline, did you ever have a major data migration or system implementation where you had several tools actually touching the data? If yes, which were the tools you touched, and was there a specific bottleneck within the process?
Mira Wise: Yeah. So one of our biggest projects over the last couple of years was actually rolling out HubSpot as a marketing automation platform. Our business used HubSpot CRM for many years, but we didn’t actually have one centralized marketing automation platform. And when you’re moving to a platform like that, not only do you need to think about how you’re going to build this for the marketing team to use, how is this going to suit the business for the next five years, how can this scale, you’re setting up all these processes from scratch, but you also have all of these different tools that you were using to essentially replicate an all-in-one system. So in our case, we were building our landing pages on WordPress, our forms were in a tool called ConvertFlow, our emails lived in Intercom. So then you’re taking all of this data and trying to consolidate it into one tool, and seeing how you can also have data continuity so there’s no line in the sand of, all that was untrackable and now we can track it via HubSpot. But also, can we pull that data in and try to create some historicals there? That can definitely be a bit of a challenge. But ultimately, moving to one centralized tool, while it’s still change management for the team, marketers need to get used to using an entirely new system, is definitely for the best of the business and makes things way more scalable to manage.
Sophie Steffen: Yeah, I can totally relate to that, because when you have different tools and you want to make sure you have one source of truth, and HubSpot is going to be the system where the truth should lie, you also want to make sure the systems playing with HubSpot are ensuring that reliability. I was wondering, was there any breaking point when you did this whole migration to HubSpot? How did you manage to ensure that the data was actually telling the truth?
Mira Wise: Yeah. Well, there were many breaking points, which makes sense for startups and scale-ups, you have legacy things that were built very early on that now you realize, oh, that was really scrappy, and now we need to adjust this because our business has grown and we’re more sophisticated. But one of the main things for us when we moved to HubSpot was, before HubSpot, we almost had a very, very simplified attribution. We basically had one field that said, this came from outbound, this came from marketing, this came from partnerships. And then when you move into HubSpot, you have more advanced attribution, which is great, but that also means now you used to have this one field that doesn’t work anymore, because now you’re looking at not just one touchpoint, but maybe first and last. And so the technical side of that was not so much the challenge. It was more getting all of the teams aligned in moving to this new model of what first and last means, and especially changing that mindset for something like a sales team, where they’re using that to pay out commissions. That’s what credit means for their team, that partnerships brought this number of leads, or sales brought this number of leads. And now you’re moving to something much more advanced. The technical side of it, of course, there were technical hurdles and many hiccups, but the most difficult part actually was the alignment, getting all of those teams on board, deciding how you’re going to track things moving forward, making sure you get their agreement, and also that they understand how it works. That kind of thing is actually more of a challenge than the technical side.
Sophie Steffen: Yeah, you mentioned attribution, which is already looking at how you want to measure success and how you want to attribute, and who gets the credit, which channel, which campaign. So looking at metrics, success metrics in this case, how do you actually know an analytics overhaul worked, and what specific metric proves to you that it was a success?
Mira Wise: Yeah. So for something like this, because you were going from a previous scenario where you really only had one data point, even though things might be a little uncomfortable moving to this new system, ultimately success ends up being trust. Now you do have these new data points, but also people understand them, you get fewer questions about what they mean. Those stakeholders know where to go and find the metrics that are important for them. All of that is more important to me than a particular KPI. It’s more, are these new dashboards you built with all of this data being used? Do people understand what they mean? Is marketing using them to actually make decisions? And when you start getting fewer of the questions like, is this right, or how does this work, can I trust this, and you see the business leaders actually using that to make decisions, then I’d consider that a success. Of course, you have some smaller things throughout, you’re going to start looking at unexpected values, you’re going to have safety nets when you’re building these things so you’re alerted when something doesn’t fit the process you defined, or something’s missing a source, to put it into a really practical example. And once you start to see those taper off because you’ve checked up all the things you missed, then it moves to trust.
Sophie Steffen: So basically your success metric is the usage of the dashboards. Do you know, for your stakeholders, the marketing team, the leadership team, is there one or two main metrics they actually use, apart from the usage obviously, to make decisions from the dashboards?
Mira Wise: Yeah. So one thing all of our go-to-market teams really focus on, because it’s really important for C-level as well, is revenue. It’s really easy to focus on high-level funnel metrics, how many leads did this campaign bring, or how many SQLs. But ideally we can actually trail that back to how much revenue came on average from that particular campaign. When they have the ability to look backwards, they become really powerful in their marketing. There could be two campaigns running, and one generates a 25% conversion rate to SQLs and one generates a 10% conversion rate. If you just stop there, you’re like, great, this 25% conversion rate campaign is actually doing much better. But if you wait to see what the revenue outcome is of those customers, and you realize that 10% is bringing us way more revenue on average per customer, then that changes your mindset a little bit in where you’re actually going to double down on your money and try to invest more. So we always try to tie it back to revenue whenever we can. Obviously that’s easier said than done, tracking is not always perfect, but when they can do that, that’s what they look at first.
Sophie Steffen: It totally makes sense to look at the actual ROI a campaign brings, because if you invest money you want to see what the results are. So for me it’s very logical that this one KPI or success metric is the one they might focus more on to make decisions on where to invest further money. But it’s interesting also to see how for you the success is actually if they use the dashboard in the first place. So moving into the hustle part, and things going south sometimes, or maybe more often than we want to admit, tell me about a real story from your experience about a reporting setup that broke, or some migration that required a massive hustle, scrambling to fix it on the fly, and how you turned that fire into a stable process.
Mira Wise: Yeah. So one practical example is when we officially made the move to HubSpot. HubSpot has a tracking pixel you put on your website, and that’s going to track all the interactions with a particular person. That’s where all of these source properties end up getting defined from, based on this cookie. And we had a lot of subdomains for our company. We had .com, .es, .fr, and we also had subdomains for where our app lives and where our signup form to create an account for our app lives. So we’d just launched HubSpot, we’d just launched our new source tracking, and we’re looking at the data and seeing that we have a really big portion of direct traffic. It’s becoming more and more common that you see that, especially with things like cookie blockers, but we realized this was actually pretty big, and we’re like, okay, we need to dive into, is something going wrong here, or is this just that now we have all of this in place, a lot of our users aren’t even giving us consent to be tracking them. And what we realized actually is that our cross-domain tracking between all of those pages was broken. So someone could go to .com and then switch to .nl, and that would be considered a whole new session, a direct conversion, a direct source, and ultimately that was not the case. So we were losing a lot of visibility into what was happening with campaigns. And from the marketing team, when we’re talking about our website, that’s within your control, your marketers can go in and fix that. But when you’re talking about something owned by the product and development team, things become much more complex, because they are typically really busy working in tight sprints. You have to book things in, and they have their own goals they’re tracking. So that created a scramble for us, because we were like, we can change this, but signups are one of the key metrics we also look at at the very top of the funnel, and if we can’t measure that correctly, we’re in a really bad place. So what we need to do is fix that immediately. And so we did a couple of things. One was working with product and engineering to make sure we were passing all that information between all of our subdomains correctly. And another part of that was actually finding a way to identify within HubSpot if something was really direct traffic, or if something was unknown traffic because cookies were declined. So there was a smart way we found with some of the HubSpot data to say, all right, of this direct traffic, let’s say 50% of it really is direct, 25% of it is from this bug that we need to fix and we’re working on, and this other 25% is actually traffic we know we can’t track. So it definitely created a scramble, but when you break it down like that toward leaders and they understand what went wrong, how we’re fixing it, and also understand, okay, it’s not quite as big, it’s 25%, not a full 100% of what we thought it was, that also helps bring the trust back, which you can lose in situations like this.
Sophie Steffen: Well, in the end, if you manage to have more accuracy, even though it needed this scramble in the beginning, now you have a better breakdown on the direct traffic, or unknown traffic you could almost say. You’re in a better place than before because now you have more accuracy in your data.
Mira Wise: Yeah, exactly. That’s a part that when you’re rolling out a system from scratch, you can always refer back to. Even if you start tracking something and the metrics aren’t positive, they’re not as good as marketing thought they would be, what you can also say is, okay, no, they’re not as good as you thought, but what this means is you now have this data at your fingertips which you didn’t have before, and you were able to figure that out within two days instead of two months. So now you can actually pivot, and you are much more data-first in that sense. So that also helps when you’re building something out from scratch and things are a little bumpy, you can say, yeah, but before we couldn’t track this at all. So we’re still moving in the right direction, even if there are some hiccups along the way.
Sophie Steffen: Yeah, and it’s always an iteration process when it comes to data, because you have new sources, new processes, new ways of measuring. So it’s always something ongoing. But in the end we’ll always have to juggle the two sides of actually doing the work, and then things that might come the way of how you can smooth those out. Moving also into the tech stack and the daily reality, before, you mentioned the tools you work on, and HubSpot not being the perfect stack but making the job most of the time. Is there a recurring limitation in your primary obstacle, which I assume is HubSpot, that is driving you completely crazy? And if there is, do you have any hack around it?
Mira Wise: Yeah. So one of my biggest qualms with HubSpot is that, my previous CRM was Salesforce, and HubSpot CRM does not have a campaign member like a Salesforce CRM has. A campaign member is basically an individual object that would track all of the touchpoints in a customer’s journey. So if they register for a webinar, you have a track of that and you know exactly where that registration came from, maybe it was from social media. Then three months later there’s another one, or they fill out a demo, and you get a really nice full version of all of those important touchpoints. And HubSpot doesn’t give you that. HubSpot gives you first and last. And that’s great, that’s still better than what we had before in our case, but from a marketing team’s perspective, you’re missing a huge chunk of visibility on what’s actually happening with everyone in your funnel. So we did a couple of hacks around this. One of those was, our data analyst actually built a custom attribution model, which lives outside of HubSpot and in Tableau, because all of those digital data points are there, it’s just not visible on HubSpot’s UI. So if you can pull that from the backend, you have the ability to actually get all of those touchpoints and build a more advanced attribution model. But what HubSpot still doesn’t do is consider all of the offline conversions that happen, even via that backend data. So let’s say someone attends an event and we upload a list of all the attendees of our session, that’s going to be missed in HubSpot’s actual tracking, because that’s not happening online. So what we actually had to do was build, essentially replicating the Salesforce campaign member inside of HubSpot, that way we could take that out and plug it into the attribution model. And you realize that changes those weights a lot, because if you’re not considering anything coming from, let’s say, referred from PartnerStack, a system our partnerships team uses to refer leads, or events, if all of that is missing and you’re giving a weight to a certain interaction, well, now you actually have five more interactions that get weights, and so all of that shifts. So that’s still probably my biggest qualm with HubSpot, that they don’t have that. But I’m happy that at least we found a solution. It’s a very advanced one, I think, but it works, and we have an attribution model that now covers everything that’s important for all of our teams, not just marketing, but also sales and partnerships. And that is a pretty good end result.
Sophie Steffen: Well, it’s pretty amazing. I think not the majority of teams would actually be able to build this workaround that plugs and plays all the systems into one to get those data, and also note to the HubSpot team, to the product backend, to develop that part, which probably would be very useful to many other teams. But yeah, good job on that, with the hack. I was pretty impressed by that.
Mira Wise: Yeah, I think so.
Sophie Steffen: Moving into the friction part, because you sit in between revenue and marketing and ops, I can imagine there is friction between the teams. Where do you see the biggest friction when you try to align the different stakeholders?
Mira Wise: Yeah. So there’s a constant back and forth between sales and marketing. Again, when we talk about, sales did this, marketing did this, and figuring out who gets credit for one thing instead of working on things together. That’s really common, that sales and marketing are constantly butting heads on this. But when it comes to the friction between operations and those teams, it’s more that oftentimes someone might come with a request, they want to track a particular data point, or they want to set up a certain automation. And in those cases they sometimes come with a solution instead of a problem. They have an idea of how it should work, and then revenue operations has to be the bearer of bad news in saying, hey, technically that’s going to be way more complicated than you think it is. And then you get into this conversation of what’s the perfect solution for right now, because we’re not gonna spend four weeks building something for an experimental campaign that we don’t know how much revenue is actually going to bring us versus something where we could put a revenue dollar on that dollar amount and say, okay, this makes sense to invest four weeks, and this is why it has this priority. So you often have to take those requests that come in and push back a little bit, understand what they’re trying to get to, and then lay out all these different options and say, this is why this might be the best solution for you.
Sophie Steffen: Yeah, and how does that work? Do you just push back, or do you have a reasoning, like priority based on estimated impact and revenue? Is there a system you have in place in order to make those decisions?
Mira Wise: Yeah. So we typically do look at the revenue that something’s going to bring. We also have a lot of projects that you’ll know are at the very top of the organization, things that C-level are really interested in and that multiple teams are driving toward. So if something falls within one of those goals, to put it into an example, right now we’re doing a lot of pushes to get our customers from monthly plans onto yearly plans. So if there’s a campaign focused around yearly conversions, we’re going to say, okay, we know this is a big company goal at the moment, we need to invest the time in this. Whereas some of these small things, you might say, does this even make sense to focus on right now, because it’s not the team’s primary KPI? So we typically align that with what the business is looking at, and then also, have we done this before, and can we prove it’s going to be worthwhile? Because what you can also do is say, all right, let’s do this in a scaled-down way for now. And if it does bring business impact, then we have a great conversation we can go have with leadership to say, we tried this out, it drives impact, now we can invest those four weeks.
Sophie Steffen: Moving into the last part of this interview, looking at the industry realities and also lessons learned the hard way. To get your totally unfiltered view about the AI topics out there, we all know that everything is flooded with AI, content, tools, analytics systems, and also changing privacy rules. So there is a change. And from your point of view, from the operational point of view, from the revenue point of view, what is your view on where B2B marketing is actually heading?
Mira Wise: Yeah. So it’s a massive, massive shift. Everything is changing within marketing, and it’s changing really quickly too, which is a difficult challenge to overcome, because you might be ahead on Friday and by Monday there’s a new model release and everything has changed and you’ve got to learn again from scratch. So it’s hard to keep up, for sure. But overall, from a marketing perspective, especially in terms of privacy and consent, I think even digital touchpoints are becoming much more difficult to measure perfectly. Absolute perfect tracking is becoming a bit unrealistic. And that’s because of privacy, but also because of AI-driven research and discovery. So all these ways we’re used to tracking things, we now need to accept a little bit of a shift toward more trends and uplift than perfect precision, which is the way we would typically measure something like an out-of-home campaign. What you might do is look at, did your traffic increase or did our revenue increase in areas where we ran these ads, because we know that’s just not as easy to track. I think we’re going to see that more and more on the digital side as well, where we know we can track a portion of this, but there’s also a portion we can’t, and then we want to look at uplift and trends over time. And so more and more often, this unknown traffic we were talking about is going to become much more common. And I think some teams are still adjusting to not having all of the perfect data points they want.
Sophie Steffen: Yeah, and it’s frustrating, because you’d think in a world that’s so digitalized that it should be possible, but at the same time with AI you can’t track back the origin, or there are just so many touchpoints. And in B2B it’s not just one touchpoint that will make the decision, in B2C also not, but in B2B usually sales cycles tend to be longer. So maybe it’s just going back to the good old manual mark on Google Analytics, do you remember that, where you just manually enter a campaign and then from that campaign you see if there’s an impact in traffic coming to the website? Maybe that, or do you think there’s something else happening with this shift?
Mira Wise: Yeah. I’m curious, because I’m sure there are tools that are going to pop up that will try to solve this. I think this is going to be something that becomes more and more common, so eventually someone’s going to say, all right, let’s think of a creative way to start measuring these things. I hope we don’t go back to manually tracking these things, but maybe we have an AI agent that does that for us, or that can analyze these types of things and give marketing teams a little bit of insight on those trends, because sometimes that’s a little more difficult to spot if you’re just looking at a dashboard. So I think we’re going to find new solutions and tools for this, but right now I think we’re still just getting into, we’re still in the acceptance phase of that, of accepting that that’s the reality. Once we get to the point of, okay, what do we do about this, I’m sure some tools are going to pop up that will help us out.
Sophie Steffen: I like the idea of an agent taking care of adding the manual timestamp on it. Looking at your personal experience, if you look back into your career, both in marketing and revenue ops, do you have any single big lesson you want to share that actually made your team more successful? Anything that really made the way you work, or your entire team, succeed?
Mira Wise: Yeah, I think there are maybe two really big ones. One of them is, when you’re working in revenue operations, really understanding your full customer lifecycle. So that goes from understanding how marketing creates their campaigns and targets and captures people, to understanding how sales converts them, to understanding how customer success actually works to grow a customer, to keep a customer. When you have all of that context, that puts you in a really good place, from being an execution-focused role of, marketing asked for this, or sales asked for this, and turns you into more of a business partner for those teams, because you get a really nice bird’s-eye view of what is happening for the full go-to-market organization, and you can also advise them on where there’s overlap or where they might be working in silos and how best to set things up. So that’s a huge one, you might not be customer facing, but know really what your customers experience end to end. And then the other point is listening, and learning how to ask really good questions, because sometimes people come with a request and it’s ultimately not what they need once you get to the core of it. So it’s really important that you dig deep into what is this person trying to accomplish, what does success really look like. Sometimes that’s a bit vague for some of the requests we get. And if you can get to the bottom of that, then that helps you build the best process or the best system or the best dashboard to give them what they need. But if you’re just kind of a yes man, and people in operations can be, because we want to help people, we want to please, so sometimes it’s like, I want to build that for you, but pushing back on that and really understanding what they’re trying to get to, and making sure you have that clearly, ultimately is going to help you build something much more suited for them in the long term.
Sophie Steffen: Yeah, that totally makes sense. It’s what you mentioned at the beginning, it’s understanding the stakeholder and what they actually want to get, what they need to make decisions. And it’s asking all those questions to understand how the end product should look. And would you also give that advice to someone who would start in your role tomorrow, or is there any additional advice you would give them?
Mira Wise: Yeah, I would definitely give them that advice. I would also give them advice on how to manage what maybe is best practice versus how your business functions, because you might have the answer in a marketing textbook of how this should work, but ultimately if that doesn’t work for your business, you’re going to need to think of a better creative solution. So it’s also what you should build versus what works for your organization. A contextual example of that is, again, our source tracking for marketing, really easy first and last. When you talk about expanding that to something like sales and partnerships, then you need to apply much more advanced business logic on this to make sure you’re tracking it properly. And I’m sure there would be some people who would say, you shouldn’t put business logic on your latest source, it should always be the most recent. But from our side, what works for us and what helps us track what we need and give the leaders the data points they need was building something that was a little bit out of the box. So finding that balance is also important.
Sophie Steffen: Yeah, this reminds me of, I wouldn’t call her a mentor, but when I was studying I was also working as a waitress in a restaurant, and she basically told me the first day, where I was doing a lot of mistakes, use your brain. So it’s basically use your brain in order to get into the role and not just execute blindly, but really think, then execute.
Mira Wise: Yeah, exactly. It’s funny that we need to have that as a reminder, but it’s an important one.
Sophie Steffen: Mira, we’re at the end of our session. Would you like to share anything else, and is there a place where people can find more about you and your work?
Mira Wise: Yeah, I’d say the best place to connect with me is probably LinkedIn. I should be more active there, I have platform fatigue, I need to get better on my LinkedIn game. But if you want to connect, you’re in Barcelona or in the Netherlands, I’m always happy to chat with people or meet up for a coffee. So feel free to shoot me a message there.
Sophie Steffen: Awesome. I’ll make sure that you’re more active on LinkedIn, I will tag you in this chat. And thank you for all those insights, Mira. To our listeners, if you found this useful, please follow us, share this video or text it to someone, and follow us. See you next time.
Herramientas mencionadas en la entrevista
Las siguientes herramientas y plataformas fueron mencionadas durante esta conversación.
